US20230085516A1 - System and method for a media intelligence platform - Google Patents

Abstract

A multi-tenant media processing platform system and method. At least a first media analysis service of a plurality of media analysis services is activated for at least a portion of an active communication session of an entity in the platform system. The first activated media analysis service performs a first media analysis on media of the active communication session that is collected by the platform system. The first activated media analysis service performs the first media analysis on the collected media while the communication session is active to generate a first media analysis result. During the active communication session, at least one media analysis result is applied.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• This application is a continuation of U.S. patent application Ser. No. 17/077,528, filed 22 Oct. 2020, which is a continuation of U.S. patent application Ser. No. 16/570,531, filed 13 Sep. 2019, which is a continuation of U.S. patent application Ser. No. 15/708,845, filed 19 Sep. 2017, which is a continuation of U.S. patent application Ser. No. 15/262,220, filed 12 Sep. 2016, which is a continuation of U.S. patent application Ser. No. 15/013,292, filed 2 Feb. 2016 which claims the benefit of U.S. Provisional Application Ser. No. 62/111,399, filed on 3 Feb. 2015, which are all incorporated in their entirety by this reference.
TECHNICAL FIELD
• This invention relates generally to the communication field, and more specifically to a new and useful system and method for a media intelligence platform in the communication field.
BACKGROUND
• There have been many changes in recent years in how people communicate. The rise of mobile devices and VoIP advances have led to numerous services and applications providing news ways for people to communicate. Audio and video processing has similarly seen continued development through the years, bringing several advances as new techniques are discovered and developed. However, it can be challenging from a development standpoint (e.g., technical expertise, time, budget, etc.) to build and maintain a communication service or application that uses advanced media processing techniques. Thus, there is a need in the communication field to create a new and useful system and method for a media intelligence platform. This invention provides such a new and useful system and method.
BRIEF DESCRIPTION OF THE FIGURES
• FIG.1 is a schematic representation of a system of an embodiment;
• FIG.2 is a flow diagram representation of a method of an embodiment;
• FIGS.3A and3B are schematic representations of a system of an embodiment;
• FIG.4 is a block diagram representation of a method of an embodiment;
• FIG.5 is a block diagram representation of a method of an embodiment;
• FIG.6 is a block diagram representation of a method of an embodiment;
• FIG.7 is a block diagram representation of a method of an embodiment;
• FIG.8 is an architecture diagram of a system of an embodiment; and
• FIG.9 is an architecture diagram of an external system.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
• The following description of preferred embodiments of the invention is not intended to limit the invention to these preferred embodiments, but rather to enable any person skilled in the art to make and use this invention.
1. System for a Media Intelligence Platform
• As shown inFIG.1, asystem100 for a media intelligence platform of a preferred embodiment includes a multitenantmedia intelligence platform101, communication resources121-123, anAPI183, anaccount configuration system130, and media processing resources141-143. Thesystem100 functions to enable media processing to be offered as a service to outside applications and services (e.g., of the app servers151 and152 ofFIG.1). The media processing is preferably used for communication-based applications. The media processing can be offered as a supplemental or complimentary feature to existing communication services (e.g., telephony call flow, call center service, customer-facing communication application, SIP trunking service, etc.). The media processing can alternatively be a stand-alone service, wherein outside communication systems can integrate with the media intelligence platform. The media intelligence platform preferably provides real-time or time-shifted analysis results, which enable communication control logic to utilize information extracted from the media content. The media intelligence platform can additionally or alternatively expose the extracted media intelligence information through communication records available through an API, a browsable user interface, an analytics portal, or any suitable interface.
• Thesystem100 can include various types of media processing resources, which can function to offer a variety of media processing techniques. The media processing techniques may include sentiment and emotion detection, transcription services, language detection, content detection, intent detection, speaker detection, and/or any suitable type of media analysis. The system can similarly be applied to one or more types of media formats.
• The media format of a communication can be text, audio, video, multi-media, or any suitable media format. The media can originate in a synchronous session or in an asynchronous message conversation. The media analysis can be performed on all or part of a communication. A communication session is preferably a synchronous media session such as a voice call, a video call, a screen-sharing call, a multi-media call, or any suitable type of media-based call. The portion of a communication can alternatively be one or more asynchronous messages. The messages can be analyzed individually, as a collection, or a time ordered sequence of messages. A message can be a text message, an image, a video, document, or any suitable media message.
• The system can additionally include a set of operational services that facilitate operation of the media analysis platform. The operational services preferably include an account system and a metering and logging service.
• Theaccount system130 of the preferred embodiment functions to map requests of the media intelligence platform to an identity within the system. The system is preferably a multitenant platform wherein multiple outside entities can create an account within the platform. An account preferably provides a set of credentials or an alternative authentication mechanism through which requests can be validated. The account system preferably authenticates requests made by an account. In one variation, an account identifier and an authentication token must be included in a request, and these authentication credentials are validated by the account system before the request is fulfilled. An account may be authenticated when making use of a REST API, when receiving signaling communication, during use of a user interface control panel or at any suitable instance.
• Various aspects of an account and usage of the platform can be configured through account management interfaces. An account may be managed through an account portal user interface. An account may alternatively be managed through API requests or through any suitable mechanism. Aspects of an account that can be managed include configuring programmatic mechanisms and settings when using the media intelligence. For example, an account manager could set various callback URIs that are triggered during errors or other events in the system. The account manager can further define various parameters that will determine how a communication is routed.
• As a related aspect, the system can include a policy engine. The policy engine may be a supplemental component or a sub-component of the account system. Policy can be set per account. Accordingly, different accounts can have different permissions, capabilities, pricing, capacity, performance, or other aspects, which can be set through an account policy. Policy may alternatively be set for a sub-account, for the entire platform, for a geographic region, or for any suitable context. Policy settings may be set by default by the platform but some or all of the policy settings may be driven by an account.
• An account can include a defined policy configuration. A policy configuration may set particular limitations for account usage within the platform. The policy can prevent an application of an outside entity incurring usage that is beyond the scope in which the application is meant to be used. For example, a policy configuration may limit the amount of media analyzed or the analysis operations performed.
• The account system can additionally include a sub-account mechanism, which functions to enable a user of the platform to partition account usage to enable multitenancy within a product of the account holder. The sub-account mechanism preferably accounts for usage, and more specifically creditable/billable usage to be monitored according to an identifier for a sub-set of usage by an account. For example, an account holder may build an outside application platform that depends on the system. The outside application platform can similarly be multitenant in that multiple users can have individually metered and accounted usage. The inheritable multitenancy property of the platform (i.e., the capability of one account within a multitenant collection of accounts to further subdivide usage the account between subclass of accounts). Can preferably provide the capabilities of a parent account to a subaccount including: billing; programmatic customization, allocation of identifiers or endpoints; and/or other customization. A billing engine can cooperatively facilitate delivering billing statements and collecting payments from the sub-accounts. Additionally, the sub-account mechanism can establish a hierarchy of policy inheritance. A platform preferably contains policies that are applied to account. In one variation, a usage policy is based on the type of account such as free account, basic account, or enterprise-account. A parent account can similarly at least partially define the policies of sub-accounts. In one variation, an account will create sub-account resources.
• The metering and logging system of the preferred embodiment functions to create a monitoring system to the media intelligence platform. The metering and logging system operates in coordination with the provided media intelligence platform resources. In one variation, the metering and logging system is integrated within an media processing resource instance running on a machine. In another variation, the metering and logging system can externally monitor the actions of the media processing resources, wherein the actions of a resource may be reported to the metering and logging system in any suitable manner. The metering and logging system functions to create a record of notable activities. The metering can be used in providing programmatic hooks (e.g., callback URI triggering, application execution, and the like), billing/crediting an associated entity (e.g., charging for services or controlling resource access), creation of an audit trail, and/or other suitable functionality. Metering preferably includes recording who (e.g., which account, sub-account, user, etc.) is associated with the use of media micro-service and the quantity of usage.
• In one variation, the metering layer of a media processing resource instance will record individual events such as media analysis initialization requests, configuration of the media analysis, size or data volume of analyzed media, changes to an analyzed media session, when a media analysis session ends, and/or any suitable event. The metering layer may additionally measure the amount of data transfer, the time of communication sessions, and/or any suitable usage metric while facilitating a particular communication session. The records are preferably metered and logged in connection to an associated account.
• A billing engine may operate independently of the metering and logging system, but may alternatively be integrated. A billing engine preferably calculates amount owed by a particular account/sub-account. The billing engine can additionally facilitate collecting and distributing of funds as appropriate. Such accounting may be used in billing or crediting an entity/account for metered usage, which functions to allow a sustainable media intelligence platform to be operated. In another variation, usage accountability can be used in limiting and balancing usage of a particular entity. As the platform is preferably multitenant, usage is preferably balanced across multiple entities. Rate limiting and action limits may be imposed at various times. Additionally, as use of a communication infrastructure is often accompanied with significant financial cost, fraudulent behavior by accounts or users of an account can be harmful to users of the platform and to the platform itself. Fraud detection can additionally be accounted for during usage of the platform.
• The system may additionally include a resource management system which functions to scale and orchestrate the computing resources that support the media intelligence platform. The set of media processing instances are preferably scaled so as to support the usage requirements across a set of accounts. As a multi-tenant platform, the resources are preferably shared across accounts. In other words, a media processing resource instance used for a first account at one time may be used by a second account at a different time. The variability of usage requirements for distinct account users is preferably normalized across a set of accounts within the platform, such that the platform is scaled to support the varying usage demands of various account holders. The resource management system can preferably instantiate more media processing resources or other services, suspend resources, or terminate service instances. In one variation, a usage model is generated for at least a set of active accounts of the platform. For the users that use the platform, or alternatively that use the media intelligence platform above a particular threshold, a model may be generated that predicts their usage over time. In one instance, an account may have a substantially steady state of usage. In another instance, the amount of usage may be a function of time of day, week, month, or year. In another instance, an account may have varying trends that are predicted in real-time based on past metrics and optionally metrics of similar accounts. As one baseline heuristic for usage prediction, the media type or application use case may be used to generate a usage model. For example, an account may select the type of media processing usage and configuration during activation—selecting a media format such as audio, video, screen sharing, and an analysis process such as sentiment and emotion detection, transcription services, language detection, content detection, intent detection, or speaker detection. A predictive model may be generated using any suitable algorithm or heuristic.
• The system may additionally include a queuing system which functions to facilitate rate limiting and/or resource management. The queuing system can preferably queue requests of a defined scope. A queuing scope may include queuing across a platform, within a regional segment of the platform, across an account, across a sub-account, across requests of a telephony endpoint, or across any suitable scope. In the variation of queuing with a platform scope, requests from different accounts to use a particular media processing resource may be initially queued until a resource is available. The requests may be queued according to different entity limits and policy. Requests of an account or sub-account may be associated with a dequeuing rate and limit. A dequeuing limit preferably defines a maximum frequency of a particular action with the media intelligence platform. A dequeuing limit preferably defines a hard limit on the number of particular actions within a time window. For example, an account may be limited to no more than a set data bandwidth of media analysis. Additionally or alternatively, the dequeuing of a request may be dependent at least in part on the resource usage and/or predicted impact of the request on the system.
2. Method for a Media Intelligence Platform
• As shown inFIG.2, amethod200 for a media intelligence platform of a preferred embodiment can include activating media analysis for at least a portion of a communication S110, collecting media for analysis S120, performing analysis on the media S130, and applying results of the analysis S140. The method functions to enable developers to extract media intelligence in a communication. The method may be applied to enable a developer to control how a communication application or service integrates with provided media analysis tools. The method can additionally be applied to offer various programmatic mechanisms for using the analysis results.
• The method can be used within a communication platform that facilitates communications. The media intelligence platform is preferably multitenant, which involves shared use of the platform from multiple and distinct entities. These entities can be managed within the platform as accounts, sub-accounts, endpoints (e.g., phone numbers, short codes, SIP addresses, or other communication addresses), and/or other entity constructs. The method preferably allows for customized treatment of media intelligence on a per entity basis. For example, a first account can use the analytic feature different from a second account. In one variation, the media analysis service can be used in a communication application platform used in defining communication logic during a call or related to a message, such as in the communication platform described in U.S. Pat. No. 8,306,021, issued 6 Nov. 2012, which is hereby incorporated in its entirety by this reference. The method may alternatively be used in combination with a communication service such as a platform that offers a customer facing communication application or any suitable communication based platform. The method can alternatively be used as a media intelligence micro-service. In a micro-service variation, the media intelligence may be the primary and possibly only service of the platform used in a communication. In one variation, negotiating and managing a communication occurs outside of the platform by a third party, and the third party system routes the communication to the platform to utilize the media intelligence capabilities of the micro-service. Routing the communication can include routing the communication through the media intelligence platform (e.g., the media intelligence platform is an intermediary node in the media path). Routing the communication can alternatively include terminating a media leg of a communication if, for example, a third party service streamed a branch of the media stream to the media intelligence platform.
• Block S110, which includes activating media analysis for at least a portion of a communication, functions to initiate or trigger use of the media analysis. Media analysis is preferably not a mandatory feature for every entity. As mentioned above, different accounts, sub-accounts, or endpoints can utilize the media intelligence independently. A portion of a communication to receive media analysis can be a full communication session, a section of a communication session. A communication session is preferably a synchronous media session such as a voice call, a video call, a screen-sharing call, a multi-media call, or any suitable type of media-based call. The portion of a communication can alternatively be one or more asynchronous messages. The messages can be analyzed individually, as a collection, or a time ordered sequence of messages.
• Activating media analysis can occur through one of various approaches. Activating media analysis may include activating media analysis in accordance with platform configuration, activating media analysis in response to a programmatic event, and/or activating media analysis in any suitable manner.
• Activating media analysis in accordance with platform configuration, functions to trigger media analysis automatically. In one variation, media analysis can be pre-configured for communications (e.g., calls or messages). Media analysis can be configured to automatically activate for communications to or from a particular endpoint, communications associated with an account or sub-account, communications made during a particular time window, communications made in association with a particular location, or any suitable communication condition. During a communication, the media intelligence platform checks if the properties of that communication map to a configuration setting that enables media analysis. The media analysis configuration can include properties of the analysis such as type of media analysis and any customized settings. In one variation, an administrator can manage media analysis configuration in a dashboard. In one variation, the method can include providing a configuration dashboard for some segment of communications and a checkbox can be used to set media analysis. There can be multiple checkboxes for different types of media analysis processes. Other suitable user-interfaces may alternatively be used. Such a configuration dashboard may enable an administrator to enable/disable media analysis for communications based on involved endpoints, sub-account, types of communications, or other categories of communications. Such media analysis configuration settings may alternatively be set through an API or any suitable interface.
• Activating media analysis in response to a programmatic event, functions to programmatically initiate and/or configure media analysis on demand. Use of programmatic events can enable the media analysis to be customized for individual communications. The programmatic event can include communication application instructions, a REST API call, a SIP message, or any suitable programmatic mechanism. The communication application instructions are preferably processed in directing state and communication flow of the communication such as with the telephony instructions of the communication platform referenced above. The REST API can alternatively be any suitable type of API call. An API call can enable media analysis to be triggered by an outside party with authentication credentials to act on that communication. The API call can provide or reference media to be analyzed. The API call can additionally configure properties of the media analysis. In a preferred embodiment, analysis can be initiated and/or configured from a cloud-based communications API. The cloud-based communications API can be used in combination with any suitable communication platform such as a telephony developers platform, a communication application platform, a call routing platform, a video chat platform, a conference call platform, a call center platform, or any suitable communications platform.
• During live media analysis, the programmatic events can similarly be used to pause, change, and/or end media analysis. Media analysis may be enabled for the full duration of a communication but may alternatively be turned on while a communication session is active. Media analysis may additionally be turned during a session. In a variation, where media analysis is performed on provided media data, the receipt of a media analysis request along with the media can act as the analysis trigger.
• Block S120, which includes collecting media for analysis, functions to obtain the media for analysis. In a preferred variation, the media intelligence platform is part of a platform that participates in routing of a media path for a communication. In synchronous communications (e.g., voice calls, video calls, multi-media streams, etc.), at least one media intelligence resource is preferably in the media path. In one variation, the platform of media intelligence may provide additional services, such as communication flow control. Such services may be used in combination with the media analysis. In another variation, a media stream may be streamed to and terminated at the media intelligence platform. For example, a third party communication system may stream a communication to the media intelligence platform, wherein the associated communication is handled through the third party communication system.
• In the case where a substantially real-time media stream is available, the method can include streaming the media through synchronous analysis resources, which functions to provide substantially real-time analysis. For example, sentiment analysis could provide a real-time analysis of a user's tone along a happiness dimension (e.g., 0 for happy 10 for angry). The real-time analysis may alternatively provide time-shifted analysis, which may be streamed analysis with a fixed or restrained delay. For example, real-time audio transcription could be provided wherein the text may be output with up to a 10 second delay. The amount of delay may be dependent on the type of analysis, the service plan level of an account, and other factors.
• In one variation, the media analysis can additionally or alternatively be performed on static media content. For example, the media content can be recorded, buffered, cached, or uploaded to the platform. Media analysis on static media content may be scheduled, queued, or managed in any suitable manner.
• The type of media analyzed in the method may include audio, video, messaging (text, image, audio, video, and/or other media), screen-sharing media, multi-media streams, and/or any suitable type of media. Preferably, the media intelligence platform can support a set of media types. The media type and the media analysis options may be dynamically determined. However, the media intelligence platform may support a single type of media. The media can be from phone calls, SIP calls, conference calls, calls with a call center system, individual text or media messages, text or media message conversations, screen sharing applications, media sharing apps, or any suitable media related application or service.
• Block S130, which includes performing analysis on media, functions to process the media and generate at least one media analysis result. At least one media analysis process is performed, but multiple media analysis processes can be used. Those processes can be performed independently, but may alternatively be dependent. Additionally, media analysis results of a first process can be used as input in a second media analysis process. For example, a transcript of an audio or video can be processed using text based sentiment analysis. The media intelligence platform can offer one or more types of media analysis processes. The performed media analysis process can have several parameters that can be defined. Media analysis parameters can be customized through the mechanism used to activate media analysis. They may alternatively be set through any programmatic mechanism.
• Additionally, the different types of media analysis processes can be updated over time. The method can include deploying a media analysis process change across multiple accounts, which functions to roll out improvements and other advances across multiple accounts. In one instance, an application using the media intelligence platform can benefit from continuous media analysis processing improvements without altering integration with the media intelligence platform.
• Some exemplary types of media analysis processes that may be performed by the media intelligence platform can include: sentiment and emotion detection, transcription, language detection, content detection, intent detection, context analysis, speaker detection, various computer vision techniques, various natural language techniques, and/or any suitable media analysis process. Some of the media analysis processes may be restricted to particular media formats. For example, there may be a first set of media analysis processes available for audio and second set of media analysis processes available for video. Different sets of media analysis processes may similarly be defined for text messages, messaging conversations, images, screen sharing media, or other types of media. Any suitable media pre-processing operations or data conditioning processes may be performed. The media analysis processes are preferably automatically applied based on the activating trigger or configuration. Media analysis processing can be performed in a remote location or some or at an endpoint device (e.g., performed by client SDK). Additionally, a portion of media analysis processing can be performed in one location and another portion at a second location. In one variation, previously trained machine learning system could be implemented in an endpoint device and media analyzed locally using the trained machine learning system. For example, an emotion detection neural network trained using millions of media streams in the cloud could be used on an endpoint device. In another variation, the analysis system located on an endpoint device could be trained using data on the device or transiting the device. This local data training variation could be used in detection algorithms. For example, a detection algorithm executed on an endpoint device could train using a facial recognition algorithm using photos in an album of the device.
• Sentiment and emotion detection functions to classify different aspects of communication. Sentiment and emotion detection can classify the tone of a communication as conveyed through the communicated content (e.g., what is said) or how the content was communicated (e.g., tone of voice). The sentiment and emotion detection can use natural language processing in extracting sentiment or emotion from content (either conveyed in text or transcribed from audio). The sentiment and emotion detection may use various voice analysis techniques for audio and video. Facial expression recognition techniques can be used in graphical media such as images and video. The sentiment emotion detection may provide overall analysis. For example, the sentiment and emotion detection may classify a call based on a measure of aggression detected throughout the call. The sentiment emotion detection may alternatively generate a time-based analysis of sentiment. For example, the measure of aggression may be a function based on time of a call. In this example, the exact moments when one of the callers was angry could be distinguished from when the callers were calm.
• Transcription functions to generate a text based machine-readable transcript of what is communicated. Various techniques in speech detection may be used. Audio and video may be transcribed into a text-based format. Transcription may additionally be used with speaker detection to assign an identity to what was said. As mentioned above, the output of a transcription service may be used in other media analysis processes such as sentiment and emotion detection.
• Language detection functions to classify the language spoken during a conversation. The language detection process may tag a communication or segments of a communication with language classifiers.
• Content detection processes function to classify topics, objects, and other content of a communication. Content detection may be similar to sentiment and emotion detection. However content detection preferably focuses on identifying, classifying, or providing classification probabilities for content in a communication. In image and video based communications, the content detection could include object detection processes. In text communications or communications with a generated transcript, topics of conversation could be classified. Tagging or classification probability assignments could be generated from content detection. The content detection can similarly be applied as an analysis of overall communication or for particular segments of a communication. Context of a communication can similarly be detected. This may be able to classify the environment of a call. For example, the context detection could identify calls made on the street, in an office, in a house, in a store, and the like.
• Intent detection functions to detect an objective of a whole communication or a segment of a communication. In one example, intent detection can be used in interpreting a request of a caller. For example, a user could ask a question and this could be parameterized into result parameterizing the question into a normalized query.
• Speaker detection functions to identify who is talking at what point in a communication. In one variation, the speaker detection marks points in time when one or more speakers are communicating. Additionally, person or entity recognition can be performed. For example, facial recognition can be used in image and video communications.
• In one variation, the method can include offering customized media processing operations. In a first variation, the method can include receiving a customized media analysis process routine. The customized media analysis process routine can be a script, a configuration document, an application, or any suitable mechanism to define media processing. The customized media analysis process routine can be executed during Block S130. The customized media analysis process routine can be statically set for a class of communications (e.g., communications for an account or subaccount, types of calls, for selected communications, etc.). The customized media analysis process routine may alternatively be retrieved from a set media process URI. The media process URI is preferably retrieved using an application layer protocol. The application layer protocol can be an HTTP-based protocol (e.g., HTTP or HTTPS), SPDY, or any suitable application layer protocol. State information of the communication can be embedded into the application layer protocol request to the media process URI. The media process URI preferably refers to a server resource managed by a third party (e.g., such as a server managed by the account holder). In one variation, third parties may be able to offer specialized or custom media analysis modules that can be enabled for use with the media analysis system for a plurality of users. Such third party media analysis modules can be metered and billed individually. Policy could be configured to set permissions of an account or endpoint to use a specific third party analysis module.
• Block S140, which includes applying results of the analysis, functions to provide access to analysis results and/or use the results. In a first variation, applying the results of the analysis includes providing access to analysis information. Access to analysis information can enable the results to be queried, accessed, and/or otherwise inspected using an API. The API is preferably a REST API. In one variation, the media analysis results are attached to communication meta-data. The media analysis results can additionally or alternatively be attached to a set of communications, wherein retrieving communication information additionally returns media analysis meta-data. A web interface could additionally provide a user interface portal that can be used to access media analysis results. The web interface could be a portal enabling search or browsing of media analysis information. Additionally or alternatively, the web interface could generate a set of infographics, graphs, tables, highlighted statistics, and other graphics that highlight analytics across a set of communications.
• The method can additionally include indexing communication sessions according to analysis results. Indexing can make the communications searchable using media analysis properties. Additionally, the indexing can enable fuzzy search based on classification probabilities. For example, a search for angry communications may return conversations ordered by highest probability of anger to lower probability of anger.
• Different types of media analysis can be applied in different ways. Transcription is preferably indexed for searchability. The transcript can additionally be linked or associated with a communication. Language detection, sentiment, content classification, context, speaker detection, and other properties can be tags associated with a communication or a particular segment of a communication.
• In some implementations, the media intelligence platform stores communication records (e.g., call records, messaging records, etc.) of a communication system (e.g., a communication system of themedia intelligence platform101, a communication system of the communication resources121-123 ofFIG.1, or a communication system external to the platform101). In some implementations, the media intelligence platform (e.g.,101) provides the stored communication records via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface. In some implementations, the media intelligence platform (e.g.,101) provides an external system (e.g., one of the app servers151,152) with the stored communication records via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface, the communication records being communication records of an account of the media intelligence platform (e.g., an account managed by the account system130) that is associated with the external system. In some implementations, the external system provides an API request (e.g., via the API183) to receive communication records for an account specified in the API request, and the media intelligence platform provides the communication records for the account to the external system via the API of the platform. In some implementations, the communication records include information for at least one communication session, and information for each communication session includes at least one of: endpoint information for each communication endpoint of the communication session, an account or sub-account associated with the communication session, endpoint device information for at least one communication endpoint of the communication session, a time period of the communication session, at least one location associated with the communication session, and media analysis results associated with the communication session. In some implementations, media analysis results included in communication records includes transcription information for a communication session, language detection information for a communication session, sentiment detection information for a communication session, emotion detection information for a communication session, content classification information for a communication session, context information for a communication session, and speaker detection information for a communication session. In some implementations, media analysis results are stored in the communication records as tags associated with a communication session (or a particular segment of a communication session). In some implementations, transcription information for a communication session (that is included in communication records, e.g., communication records for a platform account) is indexed for searchability.
• In some implementations, a request by an external system (e.g., one of the app servers151,152) for the stored communication records (e.g., provided via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface) (e.g., communication records of a media intelligence platform account of the external system) specifies media analysis results properties, and the media intelligence platform provides the external system with communication records that match the specified media analysis results properties. In some implementations, the provided communication records for each communication session include media analysis results information that matches the media analysis results properties specified in the request provided by the external system. In some implementations, the platform (e.g.,101) indexes the media analysis results information of the communication records such that the media analysis results information can be queried. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger and results of a transcription media analysis service indicates that overbilling was discussed, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger.
• An implementation of the method preferably results in the media intelligence platform observing performing media analysis on a large volume of media. The media processed is preferably used to train and improve analysis algorithms of the media intelligence platform. Analysis feedback mechanisms such as a feedback API can be used in surveying users and providing qualitative input on media analysis.
• Additionally or alternatively, applying results of the analysis can include initiating a programmatic event in response to analysis results. The programmatic event can be triggered or initiated in response to particular conditions based at least in part on the analysis results. Alternatively, the programmatic events could be initiated once a media analysis result is available, which functions to regularly act on the media analysis.
• In a first variation, the programmatic event is a selected platform event. A platform event is an action performed within the platform. The action of a platform event can include changes to a communication such as changing the media quality setting, calling an endpoint, sending an email, sending a message, activating another media analysis service, activating recording, or performing any suitable action. As one exemplary application of a platform event, all calls associated with a call center application may be monitored with emotion detection. When anger is detected, the programmatic event may turn on recording for that communication. In another exemplary situation, a customer service system could automatically connect a caller to a manager when a caller is getting too upset.
• In a second variation, the programmatic event includes transmitting a message to a configured URI, which functions to trigger a webhook or callback URI. A URI is preferably configured in association with at least one condition based on the media analysis results. When the condition is satisfied, an application layer protocol message is transmitted to the URI. The application layer protocol message preferably includes or embeds data relating to the media analysis results. An account manager may use a callback URI for integrating a system of the account manager with events in the media intelligence platform. As opposed to triggering an internal action in the media intelligence platform, the callback URI can be used to initiate any suitable process executed by an external system. In one exemplary application of a callback URI, a callback URI may be configured and then messaged when a caller is detected to be angry. Preferably the callback URI receives a POST HTTP message with information regarding the communication and/or the media analysis. The server that handles the callback URI can process the communication and media analysis information and take any suitable action.
3. Multi-Tenant Media Processing Platform System
• FIG.3A is a schematic representation of a multi-tenant mediaprocessing platform system300 that includes one or more media analysis services (e.g.,322-323).
• In some implementations, thesystem300 is similar to a system for providing a micro-services communication platform of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 is similar to thesystem100 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 is similar to a system for providing a network discovery service (e.g., STUN/TURN service) platform of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 is similar to the system1300 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 is similar to a multi-tenant media communication platform system of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 is similar to the multi-tenant media communication platform system2400 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In the implementation ofFIG.3A, the media analysis services of322 and323 are implemented as media analysis micro-services having a respective micro-service API (e.g.,392,393) and at least one media processing resource (e.g., “Entity A Resources”, “Entity C Resources”, “Entity E Resources”). In some implementations, media analysis services are implemented as media processing resources (e.g., the media processing resources ofFIG.1). In some implementations, media analysis services are implemented as media processing resource instances.
• In the embodiment ofFIG.3A, the multi-tenant mediaprocessing platform system300 includes at least one communication service (e.g.,321). In the implementation ofFIG.3A, thecommunication service321 is implemented as a communication micro-services having a respective micro-service API (e.g.,391) and at least one media processing resource (e.g., “Entity A Resources”, “Entity B Resources”, “Entity C Resources”, “Entity D Resources”). In some implementations, the communication service is implemented as a communication resource (e.g., the communication resources ofFIG.1). In some implementations, the communication service is implemented as a communication resource instance.
• Thesystem300 is a multi-tenant system that includes plural entities (e.g., entities of the systems351-355 ofFIG.3A). In some embodiments, entities include one or more of accounts, sub-accounts, organizations, users and service instances. In some implementations, each service instance includes platform configuration of theplatform system300 for an application of an account (or sub-account) of theplatform system300. For example, an account holder of a platform account can have multiple applications that use theplatform system300, each application of the account holder having a separate service instance that includes platform configuration. In some embodiments, each entity is independently configurable, and thesystem300 manages configuration for each configured entity. In some embodiments, entity configuration at thesystem300 includes configuration for one or more micro-services of thesystem300. In some embodiments, entity configuration is received from an external system of a corresponding entity via an account management interface (e.g., an account portal user interface, an account management API, and the like). In some implementations, entity configuration includes configuration similar to entity configuration of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, entity configuration includes configuration as described herein forFIGS.1 and2. In some implementations, entity configuration includes media analysis configuration, as described herein. In some implementations, entity configuration includes platform configuration, as described herein. In some implementations, entity configuration includes policy configuration, as described herein. In some implementations, entity configuration is stored at thesystem300. In some implementations, entity configuration is stored at a remote data storage device that is external to thesystem300.
• In some embodiments, thesystem300 generates one or more micro-services resources for a configured entity.
• As depicted inFIG.3A, the mediaprocessing platform system300 includes micro-services321-323,operational services399, and a media processingplatform API system383.
• Media Processing Platform
• In some embodiments, thesystem300 is a media processing platform system for multi-tenant peer-to-peer real-time media communication. In some embodiments, thesystem300 is a media processing platform system for multi-tenant peer-to-peer asynchronous media communication. In some embodiments, thesystem300 is a media processing platform system for multi-tenant peer-to-peer synchronous media communication.
• In some embodiments, each media analysis micro-service (e.g.,322,323) provides at least one media analysis service for a synchronous media stream. In some embodiments, the synchronous media stream is a synchronous media stream between two synchronous media communication endpoints (e.g., media communication endpoints of different endpoint devices). In some embodiments, the two synchronous media communication endpoints communicate via a media communication channel that is established between the two endpoints. In some embodiments, the synchronous media stream is a synchronous media stream that is broadcasted to at least one synchronous media communication endpoint. In some embodiments, a broadcasting media communication endpoint broadcasts the synchronous media stream to each destination media communication endpoint via at least one media communication channel that is established between the broadcasting endpoint and at least one destination endpoint.
• The micro-services322,323 provide at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service, as described above forFIGS.1 and2.
• In some implementations, thesystem300 includes one or more of signaling micro-services (e.g., network discovery services, such as STUN/TURN services) and media micro-services (e.g., a transcoding micro-service, a recording micro-service, a mixing micro-service, a conferencing micro-service, a media intelligence micro-service, a text-to-speech micro-service, a speech detection micro-service, a notification micro-service, a call-progress micro-service, and the like).
• In some implementations, signaling micro-services of theplatform system300 are similar to micro-services of a signaling and control system of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, signaling micro-services of theplatform system300 are similar to micro-services of the signaling andcontrol system120 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some implementations, signaling micro-services of theplatform system300 are similar to a STUN/TURN micro-service (STMS) of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, signaling micro-services of theplatform system300 are similar to the STUN/TURN micro-services (STMS) ofFIG.13 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some implementations, media analysis services of theplatform system300 are similar to media micro-services of a media service system of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, media analysis services of theplatform system300 are similar to media micro-services of themedia service system110 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, media analysis services of theplatform system300 are similar to the media micro-services1381 and1382 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some embodiments, theoperational services399 include aresource management system390, a Micro-Service (MS)interface380, anaccount system330, apolicy engine331, a metering andlogging system340, afraud detection system370, abilling engine350, and anEvent Trigger System384, as depicted inFIG.3B.
• In some implementations, theaccount system330 is similar to the account system described herein forFIG.1. In some implementations, the metering andlogging system340 is similar to the metering and logging system described herein forFIG.1. In some implementations, thepolicy engine331 is similar to the policy engine described herein forFIG.1. In some implementations, thebilling engine350 is similar to the billing engine described herein forFIG.1. In some implementations, theresource management system390 is similar to the resource management system described herein forFIG.1.
• In some implementations, thesystem300 includes a queueing system similar to the queueing system described herein forFIG.1
• In some implementations, theresource management system390, theMS interface380, theaccount system330, thepolicy engine331, the metering andlogging system340, thefraud detection system370, thebilling engine350, the communicationplatform API system383, and theEvent Trigger System384 are similar to a resource management system, a MS interface, an account system, a policy engine, a metering and logging system, a fraud detection system, a billing engine, an API service, and an Event Trigger System (respectively) of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some implementations, theresource management system390, theMS interface380, theaccount system330, thepolicy engine331, the metering andlogging system340, thefraud detection system370, thebilling engine350, the communicationplatform API system383, and theEvent Trigger System384 are similar to the resource management system1390, the MS interface1380, the account system1330, the policy engine1331, the metering and logging system1340, the fraud detection system1370, the billing engine1350, the API service1383, and the Event Trigger System1384 (respectively) ofFIG.13 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, thesystem300 includes a queueing system similar to the queueing system1360 ofFIG.13 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some implementations, one or more of the micro-services321-323, are accessible by a system (e.g., Entity Asystem351,Entity B system352,Entity C system353,Entity D system354, and Entity E system355) that is external to thesystem300 and that is a system of an entity of thesystem300. In some implementations, one or more of the micro-services321-323, are accessible by an external system of an entity via theAPI system383. In some implementations, one or more of the micro-services321-323, are accessible by an external system of an entity via a micro-service API of the respective micro-service (e.g.,micro-service A API391,micro-service B API392, micro-service C API393). In some implementations, an external system of an entity accesses one or more of the micro-services321-323 by providing at least one signaling request to the system300 (e.g., via a signaling interface of thesystem300, a queueing system, and the like).
• In some implementations, one or more of the micro-services321-323 are accessible by another micro-service of thesystem300. In some implementations, one or more of the micro-services321-323 are accessible by another micro-service of thesystem300 via theMS interface380. In some implementations, one or more of the micro-services321-323 are accessible by another micro-service of thesystem300 via theAPI service383. In some implementations, one or more of the micro-services321-323 are accessible by another micro-service of thesystem300 via a respective micro service API (e.g., one of the APIs391-393). In some implementations, one or more of the micro-services321-323 are accessible by another micro-service of thesystem300 via a signaling request.
• In some implementations, one or more of the micro-services321-323 include RESTful API resources (e.g., Entity A Resources, Entity B Resources, Entity C Resources, Entity D Resources, and Entity E Resources as depicted inFIG.3A), which act as endpoints that can act as a mechanism for specifying requested information or requesting particular actions. In some implementations, the resources are expressed as URI's or resource paths. In some implementations, the RESTful API resources are responsive to different types of HTTP methods such as GET, Put, POST and/or DELETE. In some implementations, each micro service API (e.g., one of the APIs391-393) is a RESTful API.
• In some implementations, one or more of the micro-services321-323 include a process manager, an authentication layer, and a metering layer similar to the process managers, the authentication layers, and the metering layers of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference. In some implementations, one or more of the micro-services321-323 include a process manager, an authentication layer, and a metering layer similar to the process managers, the authentication layers, and the metering layers (e.g.,1311-1313) ofFIG.13 of U.S. patent application Ser. No. 14/919,650, filed 21 Oct. 2015, which is hereby incorporated in its entirety by this reference.
• In some implementations, theplatform system300 includes media analysis service configuration for a plurality of entities (e.g., the entities corresponding to the systems351-355 ofFIG.3A) configured for use of the platform system (e.g.,300). In some implementations, the media analysis service configuration is managed by operational services of the platform system (e.g., theoperational services399 ofFIG.3A). In some implementations, the media analysis service configuration is managed by an account system of the platform system (e.g., theaccount system330 ofFIG.3B). In some implementations, the media analysis service configuration is managed by a policy engine of the platform system (e.g., thepolicy engine331 ofFIG.3B). In some implementations, the media analysis service configuration is managed by a respective micro-service of the platform system (e.g.,322-323FIG.3A).
• In some implementations, theplatform system300 includes communication service configuration for a plurality of entities (e.g., the entities corresponding to the systems351-355 ofFIG.3A) configured for use of the platform system (e.g.,300). In some implementations, the communication service configuration is managed by operational services of the platform system (e.g., theoperational services399 ofFIG.3A). In some implementations, the communication service configuration is managed by an account system of the platform system (e.g., theaccount system330 ofFIG.3B). In some implementations, the communication service configuration is managed by a policy engine of the platform system (e.g., thepolicy engine331 ofFIG.3B). In some implementations, the communication service configuration is managed by a respective micro-service of the platform system (e.g.,321FIG.3A).
4. Multi-Tenant Media Processing Platform Method
• As shown inFIG.4, themethod400 is performed at a multi-tenant media processing platform system (e.g., thesystem300 ofFIG.3A) that includes a plurality of media analysis services (e.g., implemented as micro-services322-323 ofFIG.3A).
• Themethod400 includes: activating at least a first media analysis service (e.g.,322,323) of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (process S410); the first activated media analysis service performing a first media analysis on media of the active communication session that is collected by the platform system, the first activated media analysis service performing the first media analysis on the collected media while the communication session is active to generate a first media analysis result (process S420); during the active communication session, applying at least one media analysis result (process S430).
• In some embodiments, the process S410 is similar to S110 ofFIG.2. In some embodiments, the process S420 is similar to S130 ofFIG.2. In some embodiments, the media of the active communication session is collected as described above for S120 ofFIG.2. In some embodiments, the process S430 is similar to S140 ofFIG.2.
• In some embodiments, theplatform system300 activates the first media analysis service responsive to a determination by theplatform system300 that at least one property of media of the active communication session that is collected by theplatform system300 maps to at least one setting of entity platform configuration that enables the first media analysis, the entity platform configuration being platform configuration of the entity (e.g., an entity of one of the systems351-355) in the platform system.
• In some embodiments, theplatform system300 activates the first media analysis service responsive to a communication application instruction. In some embodiments, the communication application instruction is provided by an external system of the entity (e.g., one of the systems351-355). In some embodiments, the communication application instruction is provided to thecommunication micro-service321 by an external system of the entity (e.g., one of the systems351-355), thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates the first media analysis service responsive to execution of the communication application instruction. In some embodiments, thecommunication micro-service321 receives the communication application instruction responsive to an incoming telephony communication request to an communication endpoint that is mapped to an application resource of the external system of the entity, and the communication micro-service receives the communication application instruction responsive to the communication micro-service providing an application layer request to a URI (Uniform Resource Identifier) of the external system. In some embodiments, responsive to thecommunication micro-service321 receiving an incoming telephony communication request to a communication endpoint that is mapped to an application resource of an external system (e.g., one of351-355) of the entity, the communication micro-service receives provides an application layer request to a URI (Uniform Resource Identifier) of the external system (e.g.,351-355) and receives the communication application instruction from the external system in a response to the request, thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates the first media analysis service responsive to execution of the communication application instruction by thecommunication micro-service321.
• In some embodiments, theplatform system300 activates the first media analysis service responsive to a REST API call. In some embodiments, the REST API call is provided by an external system of the entity. In some embodiments, theplatform system300 activates the first media analysis service responsive to a SIP message. In some embodiments, the SIP message call is provided by an external system of the entity.
• In some embodiments, activating at least a first media analysis service (the process S410) includes activating the first media analysis service and a second media analysis service of the plurality of media analysis services for at least a portion of the active communication session of the entity in the platform system.
• In some embodiments, theplatform system300 activates the second media analysis service responsive to a determination by theplatform system300 that at least one property of media of the active communication session that is collected by theplatform system300 maps to at least one setting of entity platform configuration that enables the second media analysis, the entity platform configuration being platform configuration of the entity (e.g., an entity of one of the systems351-355) in the platform system.
• In some embodiments, theplatform system300 activates the second media analysis service responsive to a communication application instruction. In some embodiments, the communication application instruction is provided by an external system of the entity (e.g., one of the systems351-355). In some embodiments, the communication application instruction is provided to thecommunication micro-service321 by an external system of the entity (e.g., one of the systems351-355), thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates the second media analysis service responsive to execution of the communication application instruction. In some embodiments, thecommunication micro-service321 receives the communication application instruction responsive to an incoming telephony communication request to an communication endpoint that is mapped to an application resource of the external system of the entity, and the communication micro-service receives the communication application instruction responsive to the communication micro-service providing an application layer request to a URI (Uniform Resource Identifier) of the external system. In some embodiments, responsive to thecommunication micro-service321 receiving an incoming telephony communication request to a communication endpoint that is mapped to an application resource of an external system (e.g., one of351-355) of the entity, the communication micro-service receives provides an application layer request to a URI (Uniform Resource Identifier) of the external system (e.g.,351-355) and receives the communication application instruction from the external system in a response to the request, thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates the second media analysis service responsive to execution of the communication application instruction by thecommunication micro-service321.
• In some embodiments, theplatform system300 activates the second media analysis service responsive to a REST API call. In some embodiments, the REST API call is provided by an external system of the entity. In some embodiments, theplatform system300 activates the second media analysis service responsive to a SIP message. In some embodiments, the SIP message call is provided by an external system of the entity.
• In some embodiments, theplatform system300 activates each media analysis service responsive to at least one of: a determination by the platform system that at least one property of media of the active communication session that is collected by the platform system maps to at least one setting of entity platform configuration that enables media analysis, the entity platform configuration being platform configuration of the entity in the platform system; a communication application instruction; a REST API call; and a SIP message. In some embodiments, the communication application instruction is provided by an external system of the entity (e.g., one of the systems351-355). In some embodiments, the communication application instruction is provided to thecommunication micro-service321 by an external system of the entity (e.g., one of the systems351-355), thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates a media analysis service responsive to execution of the communication application instruction. In some embodiments, thecommunication micro-service321 receives the communication application instruction responsive to an incoming telephony communication request to an communication endpoint that is mapped to an application resource of the external system of the entity, and the communication micro-service receives the communication application instruction responsive to the communication micro-service providing an application layer request to a URI (Uniform Resource Identifier) of the external system. In some embodiments, responsive to thecommunication micro-service321 receiving an incoming telephony communication request to a communication endpoint that is mapped to an application resource of an external system (e.g., one of351-355) of the entity, the communication micro-service receives provides an application layer request to a URI (Uniform Resource Identifier) of the external system (e.g.,351-355) and receives the communication application instruction from the external system in a response to the request, thecommunication micro-service321 executes the communication application instruction, and theplatform system300 activates a media analysis service responsive to execution of the communication application instruction by thecommunication micro-service321
• In some embodiments, the entity platform configuration is provided by at least one external system (e.g.,351-355) via an account management interface (e.g., of theAPI System383 ofFIG.3) of the mediaprocessing platform system300, the account management interface including a least one of an account portal user interface and an account management API.
• In some embodiments in which the first media analysis service and the second media analysis service are activated, the first media analysis service and the second media analysis service are both active during at least a portion of the active communication session.
• In some embodiments, the platform configuration specifies automatic activation of at least one media analysis service for at least one of: a communication of the entity to a particular endpoint; a communication session of the entity from a particular endpoint; a communication session of the entity associated with a particular account; a communication session of the entity made during a particular time window; and a communication session of the entity made in association with a particular location.
• In some embodiments, communication application instructions are processed in directing state and communication flow of the communication session.
• Performing Media Analysis
• In some embodiments, themethod400 includes activating a second media analysis service of the plurality of media analysis services for at least a portion of the active communication session of the entity in the platform system, and the process S420 includes the second activated media analysis service performing a second media analysis on the collected media while the communication session is active to generate a second media analysis result, the second media analysis being different from the first media analysis.
• In some embodiments, activating at least a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (process S410) includes: theplatform system300 selecting at least the first media analysis service from the plurality of media analysis services of theplatform system300, the plurality of media analysis services including at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service; and theplatform system300 activating each selected media analysis service. In some embodiments, activating at least a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (process S410) includes: theplatform system300 selecting at least the first media analysis service from the plurality of media analysis services of theplatform system300, the plurality of media analysis services including at least one of a sentiment and emotion detection service, a context analysis service, and a computer vision service; and theplatform system300 activating each selected media analysis service. In some implementations, theplatform system300 selects each selected media analysis service responsive to a programmatic event. In some implementations, the programmatic event is provided by an external system (e.g.,351-355 ofFIG.3) of the entity. In some implementations, the programmatic event includes reception at theplatform system300 of a communication application instruction that initiates the activation. In some implementations, the programmatic event includes reception at theplatform system300 of a REST API call that initiates the activation. In some implementations, theplatform system300 selects each selected media analysis service in accordance with entity platform configuration of the entity.
• In some embodiments, activating at least a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (process S410) includes: activating a sentiment and emotion detection service.
• In some embodiments, the first media analysis service collects the media of the active communication session. In some embodiments, the second media analysis service collects the media of the active communication session.
• In some embodiments, a REST API call that activates a media analysis service is provided by an outside party with authentication credentials to act on the communication session, the API call providing the media of the communication session to be analyzed and configuration of the media analysis service.
• In some embodiments, the process S410 includes activating a second media analysis service of the plurality of media analysis services for at least a portion of the active communication session of the entity in the platform system, and themethod400 includes the second media analysis service performing a second media analysis on the first media analysis result while the communication session is active to generate a second media analysis result, the second media analysis being different from the first media analysis. In some implementations, the first media analysis service transmits an application layer protocol message to a Uniform Resource Identifier (URI) of the second media analysis service, the application layer protocol message providing data relating to the first media analysis result. In some embodiments, a callback URI parameter of the first media analysis service is configured to specify the URI of the second media analysis service.
• In some implementations, the media is real-time media, and the first media analysis service performs real-time media analysis on the real-time media. In some implementations, the media is stored media, and the first media analysis service performs asynchronous media analysis on the stored media.
• In some implementations in which a second media analysis service is activated, the media is real-time media, and the second media analysis service performs real-time media analysis on the real-time media. In some implementations in which a second media analysis service is activated, the media is stored media, and the second media analysis service performs asynchronous media analysis on the stored media.
• In some implementations, the first media analysis service is a sentiment and emotion detection service. In some implementations in which a second media analysis service is activated, the second media analysis service is a sentiment and emotion detection service, and the first media analysis service is a different media analysis service. In some implementations in which a second media analysis service is activated, the first media analysis service is a sentiment and emotion detection service, and the second media analysis service is a different media analysis service.
• Collecting Media for Analysis
• In some implementations, the mediaprocessing platform system300 collects media of the active communication session by using a communication service of the system300 (e.g., the communication service provided by themicro-service321 ofFIG.3). In some implementations, the mediaprocessing platform system300 collects media of the active communication session from an external system, such as, for example, a communication system that is external to the mediaprocessing platform system300, a communication endpoint device (e.g.,356,357 ofFIG.3), a system of the entity (e.g., one of the systems351-355 ofFIG.3). In some implementations, a communication endpoint device includes at least one of a mobile phone, computer, tablet, portable computing device, and the like.
• In some implementations, the first activated media analysis service (e.g., the media analysis service of the micro-service322) collects media of the active communication session from a communication service of the system300 (e.g., the communication service of themicro-service321 ofFIG.3). In some implementations, the first activated media analysis service (e.g., the media analysis service of the micro-service322) collects media of the active communication session from another media analysis service of the system300 (e.g., a media analysis service of themicro-service323 ofFIG.3). In some implementations, the first activated media analysis service collects media of the active communication session from an external system, such as, for example, a communication system that is external to the mediaprocessing platform system300, a communication endpoint device (e.g.,356,357 ofFIG.3), a system of the entity (e.g., one of the systems351-355 ofFIG.3). In some implementations, a communication endpoint device includes at least one of a mobile phone, computer, tablet, portable computing device, and the like.
• In some implementations, a second activated media analysis service (e.g., the media analysis service of the micro-service323) collects media of the active communication session from a communication service of the system300 (e.g., the communication service of themicro-service321 ofFIG.3). In some implementations, the second activated media analysis service (e.g., the media analysis service of the micro-service323) collects media of the active communication session from another media analysis service of the system300 (e.g., a media analysis service of themicro-service322 ofFIG.3). In some implementations, the second activated media analysis service collects media of the active communication session from an external system, such as, for example, a communication system that is external to the mediaprocessing platform system300, a communication endpoint device (e.g.,356,357 ofFIG.3), a system of the entity (e.g., one of the systems351-355 ofFIG.3). In some implementations, a communication endpoint device includes at least one of a mobile phone, computer, tablet, portable computing device, and the like.
• Applying Media Analysis Results
• In some embodiments, applying at least one media analysis result includes applying the first media analysis result. In some implementations in which the first media analysis service and the second media analysis service are activated, applying at least one media analysis result includes applying the first media analysis result. In some implementations in which the first media analysis service and the second media analysis service are activated, applying at least one media analysis result includes applying the second media analysis result. In some implementations in which the first media analysis service and the second media analysis service are activated, applying at least one media analysis result includes applying the first media analysis result and the second media analysis result.
• Applying Media Analysis Results: Changing the Active Communication Session
• In some embodiments, applying the at least one media analysis result (process S430) includes changing the active communication session in response to the first media analysis result.
• In some embodiments, applying the at least one media analysis result (process S430) includes transmitting at least one application layer protocol message to at least one callback URI of a communication service of the platform system (e.g., a callback URI of thecommunication micro-service321, such as, for example, a callback URI corresponding to the communication micro-service API391) of the entity, the at least one application layer protocol message providing the at least one media analysis result.
• In some embodiments in which the first media analysis service and the second media analysis service are activated, applying the at least one media analysis result (process S430) includes changing the active communication session in response to the first media analysis result and the second media analysis result. In some embodiments in which the first media analysis service and the second media analysis service are activated, applying the at least one media analysis result (process S430) includes changing the active communication session in response to the second media analysis result. In some embodiments in which the first media analysis service and the second media analysis service are activated, applying the at least one media analysis result (process S430) includes changing the active communication session in response to the first media analysis result.
• In some implementations, changing the active communication session includes changing media of the communication session.
• In some implementations, changing the active communication session includes at least one of: changing a media quality of the active communication session; calling an endpoint, sending an e-mail, sending a message, activating another media analysis service, and activating recording. In some implementations, the communication service (e.g., of321) changes the active communication session.
• Applying Media Analysis Results: Providing Analysis Results Using a Callback URI
• In some embodiments, applying the at least one media analysis result (process S430) includes transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by entity platform configuration. In some implementations, the at least one application layer protocol message provides the first media analysis result.
• In some implementations in which the first media analysis service and the second media analysis service are activated, the at least one application layer protocol message provides the first media analysis result. In some implementations in which the first media analysis service and the second media analysis service are activated, the at least one application layer protocol message provides the second media analysis result. In some implementations in which the first media analysis service and the second media analysis service are activated, the at least one application layer protocol message provides the first media analysis result and the second media analysis result.
• In some embodiments, applying the at least one media analysis result (process S430) includes applying the first media analysis result by transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by a REST API call that activates the first media analysis service, the at least one application layer protocol message providing the first media analysis result.
• Applying Media Analysis Results: Providing Access
• In some embodiments, applying the at least one media analysis result (process S430) includes: providing access to the at least one media analysis result as described above for S140. In some embodiments, applying the at least one media analysis result (process S430) includes providing access to the at least one media analysis result via a REST API (e.g., the an API of the API system383). In some embodiments, applying the at least one media analysis result (process S430) includes: providing access to the at least one media analysis result by attaching the at least one media analysis result to communication meta-data (e.g., communication metadata provided by thecommunication service321 ofFIG.3). In some embodiments, applying the at least one media analysis result (process S430) includes providing access to the at least one media analysis result via a user interface portal provided by thesystem300. In some embodiments, thesystem300 indexes communication sessions (e.g., communication sessions of the communication micro-service of321) according to the at least one media analysis results.
• In some implementations, providing access to the at least one media analysis result includes: indexing communication sessions according to analysis results. Indexing can make the communications searchable using media analysis properties. Additionally, the indexing can enable fuzzy search based on classification probabilities. For example, a search for angry communications may return conversations ordered by highest probability of anger to lower probability of anger. Different types of media analysis can be applied in different ways. Transcription is preferably indexed for searchability. The transcript can additionally be linked or associated with a communication. Language detection, sentiment, content classification, context, speaker detection, and other properties can be tags associated with a communication or a particular segment of a communication.
• In some implementations, providing access to the at least one media analysis result includes: theplatform system300 providing stored communication records via at least one of an API (e.g.,383 ofFIG.3), a browsable user interface, an analytics portal, or any suitable interface. In some implementations, theplatform system300 stores communication records (e.g., call records, messaging records, etc.) of a communication system (e.g., a communication system of theplatform system300, a communication system of the micro-service321, a communication system external to the system300). In some implementations, thesystem300 provides an external system (e.g.,351-355) with the stored communication records via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface, the communication records being communication records of an account of the platform system (e.g., an account managed by the account system330) that is associated with the external system. In some implementations, the external system provides an API request (e.g., via the API383) to receive communication records for an account specified in the API request, and the platform system provides the communication records for the account to the external system via the API of the platform.
• In some implementations, the communication records include information for at least one communication session, and information for each communication session includes at least one of: endpoint information for each communication endpoint of the communication session, an account or sub-account associated with the communication session, endpoint device information (e.g., of adevice356,357 ofFIG.3) for at least one communication endpoint of the communication session, a time period of the communication session, at least one location associated with the communication session, and media analysis results associated with the communication session. In some implementations, media analysis results included in communication records includes the at least one media analysis result. In some implementations, media analysis results included in communication records includes transcription information for a communication session, language detection information for a communication session, sentiment detection information for a communication session, emotion detection information for a communication session, content classification information for a communication session, context information for a communication session, and speaker detection information for a communication session. In some implementations, media analysis results are stored in the communication records as tags associated with a communication session (or a particular segment of a communication session). In some implementations, transcription information for a communication session (that is included in communication records, e.g., communication records for a platform account) is indexed for searchability.
• In some implementations, a request by an external system (e.g., one of the app servers351-355) for the stored communication records (e.g., provided via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface) (e.g., communication records of a platform system account of the external system) specifies media analysis results properties, and the platform system provides the external system with communication records that match the specified media analysis results properties. In some implementations, the provided communication records for each communication session include media analysis results information that matches the media analysis results properties specified in the request provided by the external system. In some implementations, the platform system (e.g.,300) indexes the media analysis results information of the communication records such that the media analysis results information can be queried. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger and results of a transcription media analysis service indicates that overbilling was discussed, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger.
• Accounting and Billing
• In some embodiments, themethod400 includes: accounting for the use of the first activated media analysis service on behalf of the entity (process S440); and generating billing information for the entity based on the accounting for the use of the first activated media analysis service (process S450).
• In some embodiments in which the first media analysis service and the second media analysis service are activated, themethod400 includes: accounting for the use of the first activated media analysis service on behalf of the entity (process S440); accounting for the use of the second activated media analysis service on behalf of the entity, the platform system accounting for the use of the second activated media analysis service independently from the accounting for the use of the first activated media analysis service (process S440); and generating billing information for the entity based on the accounting for the use of the first activated media analysis service and the accounting for the use of the second activated media analysis service (process S450).
Additional Embodiments
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service.
• In some embodiments, the active communication session is a communication session between at least a first communication endpoint device (e.g.,356) and a second communication endpoint device (e.g.,357 ofFIG.3), the first communication endpoint device and the second communication endpoint device each being external to theplatform system300, and each media analysis service (e.g., each service of the media analysis micro-services322-323) is constructed to perform media analysis on media of the communication session between at least the first communication endpoint device and the second communication endpoint device.
• In some embodiments, plurality of media analysis services includes at least one of a sentiment and emotion detection service, a language detection service, an intent detection service, a speaker detection service, a context analysis service, and a computer vision service.
• In some embodiments, the active communication session is a peer-to-peer communication session.
• In some embodiments, the active communication session is a real-time communication session.
• In some embodiments, the media is a real-time media stream.
• In some embodiments, at least one REST API call is provided by an external system of the entity.
• In some embodiments, the first media analysis service includes a customized media analysis process routine, as described above for block S130.
• In some embodiments, themethod400 includes: thesystem300 training at least one media analysis service by using the first media analysis result. In some embodiments, themethod400 includes: thesystem300 training at least one media analysis service by using the first media analysis result and user feedback provided by an external system (e.g.,351-357) via a feedback API of the system300 (e.g., a feedback API of the API system383). In some implementations, thesystem300 receives user feedback via the feedback API, the feedback indicating a quality of media analysis provided by the first media analysis service, and thesystem300 trains the first media analysis service based on the received feedback. In some implementations, the system maintains a history data that includes media analysis results and corresponding media and user feedback of the first media analysis service, and the system trains the first media analysis service by using the history data.
• Method Operation
• In some implementations, the process S410 is performed by theAPI system383. In some implementations, the process S410 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S410 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S410 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S410 is performed by theoperational services399. In some implementations, the process S410 is performed by theaccount system330. In some implementations, the process S410 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S410 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S420 is performed by at least one media analysis micro-service (e.g.,322,323). In some implementations, the process S420 is performed by theAPI system383 and at least one media analysis micro-service. In some implementations, the process S420 is performed by theAPI system383, at least one media analysis micro-service, and the operations services399.
• In some implementations, the process S430 is performed by theAPI system383. In some implementations, the process S430 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S430 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S430 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S430 is performed by theoperational services399. In some implementations, the process S430 is performed by theaccount system330. In some implementations, the process S430 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S430 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S440 is performed by theAPI system383. In some implementations, the process S440 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S440 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S440 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S440 is performed by theoperational services399. In some implementations, the process S440 is performed by theaccount system330. In some implementations, the process S440 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S440 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S450 is performed by theAPI system383. In some implementations, the process S450 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S450 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S450 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S450 is performed by theoperational services399. In some implementations, the process S450 is performed by theaccount system330. In some implementations, the process S450 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S450 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
5. Multi-Tenant MediaProcessing Platform Method500
• As shown inFIG.5, themethod500 is performed at a multi-tenant media processing platform system (e.g., thesystem300 ofFIG.3A) that includes a plurality of media analysis services (e.g., implemented as micro-services322-323 ofFIG.3A).
• The method500 includes: activating a first media analysis service (e.g., of the micro-services322,323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session of an entity (e.g., an entity of one of systems351-355) in the platform system, the platform system activating the first media analysis service responsive to a REST API call provided by an external system (e.g.,351-355) of the entity (process S510); the first activated media analysis service performing a first media analysis on media of the active communication session that is collected by the first activated media analysis service, while the communication session is active, to generate a first media analysis result (process S520); during the active communication session, applying the first media analysis result by transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by the REST API call, the at least one application layer protocol message providing the first media analysis result (process S530); accounting for the use of the first activated media analysis service on behalf of the entity (process S540); generating billing information for the entity based on the accounting for the use of the first activated media analysis service (process S550). Each external system (e.g.,351-355) of the entity is a system of an account holder of a platform account of the multi-tenant media processing platform system, each external system being external to theplatform system300.
• Selection of at Least One Media Analysis Service
• In some embodiments, activating a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (the process S510) includes: theplatform system300 selecting at least the first media analysis service from the plurality of media analysis services of theplatform system300, the plurality of media analysis services including at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service; and theplatform system300 activating each selected media analysis service. In some embodiments, activating a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (the process S510) includes: theplatform system300 selecting at least the first media analysis service from the plurality of media analysis services of theplatform system300, the plurality of media analysis services including at least one of a sentiment and emotion detection service, a context analysis service, and a computer vision service; and theplatform system300 activating each selected media analysis service. In some implementations, theplatform system300 selects each selected media analysis service responsive to a programmatic event. In some implementations, the programmatic event is provided by an external system (e.g.,351-355 ofFIG.3) of the entity. In some implementations, the programmatic event includes reception at theplatform system300 of a communication application instruction that initiates the activation. In some implementations, the programmatic event includes reception at theplatform system300 of a REST API call that initiates the activation. In some implementations, theplatform system300 selects each selected media analysis service in accordance with entity platform configuration of the entity.
• In some embodiments, activating at least a first media analysis service of the plurality of media analysis services for at least a portion of an active communication session of an entity in the platform system (process S410) includes: activating a sentiment and emotion detection service.
• Activation REST API Call and Callback URI
• In some implementations, the first media analysis service is activated for a specified active communication session, such as for example, a pre-established communication session that is active prior to activation of the first media analysis service. In some implementations, the first media analysis service is activated for all active communication sessions of the entity.
• In some implementations, the activation REST API call to activate the first media analysis service specifies at least the communication session and at least the first activated media analysis service. In some implementations, the REST API call to activate the first media analysis service specifies at least the communication session, at least the first activated media analysis service, and at least the callback URI for the first media analysis service. In some implementations, the REST API call to activate the first media analysis service specifies at least the communication session, at least the first activated media analysis service, and at least media analysis service configuration for the first media analysis service. In some implementations, the activation REST API call specifies the communication session by using a session ID of a communication session of the communication micro-service (e.g.,321) of theplatform300.
• In some implementations, the activation REST API call to activate the first media analysis service includes (or references) the media of the communication session, and specifies at least the first activated media analysis service. In some implementations, the REST API call to activate the first media analysis service includes (or references) the media of the communication session, and specifies at least the first activated media analysis service, and at least the callback URI for the first media analysis service. In some implementations, the REST API call to activate the first media analysis service includes (or references) the media of the communication session, and specifies at least the first activated media analysis service, and at least media analysis service configuration for the first media analysis service.
• In some implementations, the callback URI is specified by a callback URI parameter of the media analysis service configuration of the activation REST API call. In some implementations, the callback URI is specified by a callback URI parameter of the media analysis service configuration of the first media analysis micro-service (e.g.,322,323). In some implementations, the callback URI is specified by a callback URI parameter of the media analysis service configuration of theaccount system330.
• In some implementations, the callback URI is specified by a callback URI parameter included in entity platform configuration of the entity (e.g., entity platform configuration of theaccount system330 ofFIG.3).
• In some implementations, the callback URI for the first media analysis service is provided by at least one external system (e.g.,351-355) of the entity via an account management interface (e.g., of theAPI System383 ofFIG.3) of the mediaprocessing platform system300, the account management interface including a least one of an account portal user interface and an account management API.
• Activation REST API Call and Callback URI: Two Media Analysis Services
• In some implementations, the REST API specifies: the first activated media analysis service, a second media analysis service, the callback URI for the first media analysis result, and a callback URI for a second media analysis result generated by the second media analysis service. In some implementations, the REST API specifies the first activated media analysis service, a second media analysis service, and the callback URI, and the specified callback URI is for the first media analysis result and a second media analysis result generated by the second media analysis service.
• In some implementations, the REST API specifies: the first activated media analysis service, a second media analysis service, the callback URI for the first media analysis result, and a platform event to be performed based on the second media analysis result generated by the second media analysis service.
• In some implantations, the platform event is an action performed within theplatform300. In some implementations, the platform event is an action performed by a communication micro-service (e.g.,321) of theplatform300. In some implementations, action of a platform event can include changes to a communication such as changing the media quality setting, calling an endpoint, sending an email, sending a message, activating another media analysis service, activating recording, or performing any suitable action. In some implementations, an exemplary application of a platform event includes monitoring all calls associated with a call center application with emotion detection. In some implementations, an exemplary application of a platform event includes monitoring all calls associated with a call center application with emotion detection; when anger is detected, the programmatic event turns on recording for that communication. In some implementations, an exemplary application of a platform event includes a customer service system automatically connecting a caller to a manager when a caller is getting too upset.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: activating the second media analysis service (e.g., of the micro-services322,323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session of an entity (e.g., an entity of one of systems351-355) in the platform system, the platform system activating the second media analysis service responsive to a REST API call provided by an external system (e.g.,351-355) of the entity.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: the second activated media analysis service performing a second media analysis on media of the active communication session that is collected by the second activated media analysis service, while the communication session is active, to generate a second media analysis result.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: the second activated media analysis service performing a second media analysis on the first analysis result to generate a second media analysis result.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: during the active communication session, applying the second media analysis result by transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by the REST API call to activate the second media analysis service, the at least one application layer protocol message providing the second media analysis result.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: during the active communication session, applying the second media analysis result by performing a platform event, as described herein for S140 ofFIG.1.
• In some embodiments in which the second media analysis service is activated, themethod500 includes: accounting for the use of the second activated media analysis service on behalf of the entity; and generating billing information for the entity based on the accounting for the use of the second activated media analysis service.
• In some embodiments in which the second media analysis service is activated, the first media analysis service and the second media analysis service are both active during at least a portion of the active communication session.
• In some embodiments in which the second media analysis service is activated, use of the second activated media analysis service is accounted for independently from use of the first activated media analysis service.
• In some embodiments in which the second media analysis service is activated, media analysis of the second media analysis service is different from media analysis of the first media analysis service.
• In some embodiments in which the second media analysis service is activated, a callback URI parameter of the first media analysis service specifies a URI of the second media analysis service, the first media analysis service transmits an application layer protocol message to the URI of the second media analysis service, the message providing data relating to the first media analysis result, the second media analysis service performing a second media analysis on the first media analysis result while the communication session is active to generate a second media analysis result, the second media analysis result being different from the first media analysis result.
• In some implementations, the media is real-time media, and the first media analysis service performs real-time media analysis on the real-time media. In some implementations, the media is stored media, and the first media analysis service performs asynchronous media analysis on the stored media.
• In some embodiments in which the second media analysis service is activated, the media is real-time media, and the first media analysis service performs real-time media analysis on the real-time media. In some embodiments in which the second media analysis service is activated, the media is stored media, and the first media analysis service performs asynchronous media analysis on the stored media.
• In some embodiments in which the second media analysis service is activated, the media is real-time media, and the second media analysis service performs real-time media analysis on the real-time media. In some embodiments in which the second media analysis service is activated, the media is stored media, and the second media analysis service performs asynchronous media analysis on the stored media.
• In some implementations, the first media analysis service is a sentiment and emotion detection service. In some implementations in which a second media analysis service is activated, the second media analysis service is a sentiment and emotion detection service, and the first media analysis service is a different media analysis service. In some implementations in which a second media analysis service is activated, the first media analysis service is a sentiment and emotion detection service, and the second media analysis service is a different media analysis service.
• Deactivating Media Analysis
• In some embodiments, themethod500 includes: de-activating the first media analysis service (e.g., of the micro-services322,323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session of an entity (e.g., an entity of one of systems351-355) in the platform system, the platform system de-activating the first media analysis service responsive to a de-activation REST API call provided by the external system. In some implementations, the REST API call to de-activate the first media analysis service specifies at least the communication session and at least the first activated media analysis service. In some implementations, the de-activation REST API call specifies the communication session by using a session ID of a communication session of the communication micro-service (e.g.,321) of theplatform300. In some implementations, the REST API call to de-activate the first media analysis service specifies at least the first activated media analysis service. In some implementations, the REST API call to de-activate the first media analysis service specifies at least the first activated media analysis service and a media identifier.
• In some implementations, the first media analysis service is de-activated for a specified active communication session. In some implementations, the first media analysis service is de-activated for all active communication sessions of the entity.
• Configuring Media Analysis
• In some embodiments, themethod500 includes: configuring the first media analysis service (e.g., of the micro-services322,323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session of an entity (e.g., an entity of one of systems351-355) in the platform system, the platform system configuring the first media analysis service responsive to a configuration REST API call provided by the external system. In some implementations, the REST API call to configure the first media analysis service specifies at least the first activated media analysis service, and specifies configuration for at least the first activated media analysis service. In some implementations, the REST API call to configure the first media analysis service specifies the active communication session, specifies at least the first activated media analysis service, and specifies configuration for at least the first activated media analysis service. In some implementations, configuration of the configuration REST API call includes the callback URI of the first media analysis service.
• In some implementations, the activation REST API call specifies the communication session by using a session ID of a communication session of the communication micro-service (e.g.,321) of theplatform300.
• In some implementations, the REST API call to configure the first media analysis service specifies at least the first activated media analysis service and specifies configuration for at least the first activated media analysis service. In some implementations, the REST API call to configure the first media analysis service specifies at least the first activated media analysis service, and specifies configuration for at least the first activated media analysis service, and a media identifier.
• In some implementations, the first media analysis service is configured for a specified active communication session, such as for example, a pre-established communication session that is active prior to configuration of the first media analysis service. In some implementations, the first media analysis service is configured for all active communication sessions of the entity.
• Customized Media Analysis
• In some embodiments, the activation REST API specifies a customized media analysis service (that includes a customized media analysis process routine) as the first media analysis service, as described above for block S130. In some embodiments, the activation REST API includes program instructions (or a link to program instructions) for the customized media analysis service, and responsive to the activation REST API request, theplatform system300 generates a customized media analysis micro-service (e.g.,322,323) that provides a media analysis service according to the received program instructions.
• In some embodiments, the activation REST API includes program instructions (or a link to program instructions) for the customized media analysis service, and responsive to the activation REST API request, theplatform system300 controls a communication micro-service (e.g.,321) to provide a media analysis service according to the received program instructions.
• APIs
• In some implementations, the REST API call is an API call of thePlatform API System383. In some implementations, the REST API call is an API call of the first media analysis service (e.g., an API call of one of the mediaanalysis micro-service APIs392,393). In some implementations, the REST API call is an API call of a communication service of the platform (e.g., an API call of the communication micro-service API391).
• In some implementations, an activation REST API call is an API call of thePlatform API System383. In some implementations, an activation REST API call is an API call of a respective media analysis service (e.g., an API call of one of the mediaanalysis micro-service APIs392,393). In some implementations, an activation REST API call is an API call of a communication service of the platform (e.g., an API call of the communication micro-service API391).
• In some implementations, a de-activation REST API call is an API call of thePlatform API System383. In some implementations, a de-activation REST API call is an API call of a respective media analysis service (e.g., an API call of one of the mediaanalysis micro-service APIs392,393). In some implementations, a de-activation REST API call is an API call of a communication service of the platform (e.g., an API call of the communication micro-service API391).
• In some implementations, a configuration REST API call is an API call of thePlatform API System383. In some implementations, a configuration REST API call is an API call of a respective media analysis service (e.g., an API call of one of the mediaanalysis micro-service APIs392,393). In some implementations, a configuration REST API call is an API call of a communication service of the platform (e.g., an API call of the communication micro-service API391).
• Stand-Alone Media Analysis
• In some implementations, the activation REST API call to activate the first media analysis service includes (or references) the media of the communication session, specifies at least the first activated media analysis service, and the activation REST API call is an API all of the first media analysis service (e.g., an API call of one of the mediaanalysis micro-service APIs392,393). For example, as shown inFIG.3, in an exemplary API call, the “ENTITY E System”355 provides an activation REST API call to the mediaanalysis micro-service C323 via themicro-service's API393, and the activation REST API call provides the media to be analyzed by the media analysis service of themicro-service323. In such an example, the media can be media of a communication session that is managed by a communication platform (or service) that is external to theplatform system300. In such an example, the micro-service323 can be implemented as a sentiment and emotion detection media analysis micro-service.
• Feedback
• In some embodiments, themethod500 includes: thesystem300 training at least one media analysis service by using the first media analysis result. In some embodiments, themethod500 includes: thesystem300 training at least one media analysis service by using the first media analysis result and user feedback provided by an external system (e.g.,351-357) via a feedback API of the system300 (e.g., a feedback API of the API system383). In some implementations, thesystem300 receives user feedback via the feedback API, the feedback indicating a quality of media analysis provided by the first media analysis service, and thesystem300 trains the first media analysis service based on the received feedback. In some implementations, the system maintains a history data that includes media analysis results and corresponding media and user feedback of the first media analysis service, and the system trains the first media analysis service by using the history data.
• Applying Media Analysis Results to a Communication Service
• In some embodiments, themethod500 includes: during the active communication session, applying the first media analysis result by transmitting at least one application layer protocol message to at least one callback URI of a communication service of the platform system (e.g., a callback URI of thecommunication micro-service321, such as, for example, a callback URI corresponding to the communication micro-service API391) of the entity, the at least one application layer protocol message providing the first media analysis result.
• Applying Media Analysis: Providing Access to Results
• In some embodiments, themethod500 includes: providing access to the first media analysis result as described above for S140.
• In some implementations, providing access to the first media analysis result includes: indexing communication sessions according to analysis results. Indexing can make the communications searchable using media analysis properties. Additionally, the indexing can enable fuzzy search based on classification probabilities. For example, a search for angry communications may return conversations ordered by highest probability of anger to lower probability of anger. Different types of media analysis can be applied in different ways. Transcription is preferably indexed for searchability. The transcript can additionally be linked or associated with a communication. Language detection, sentiment, content classification, context, speaker detection, and other properties can be tags associated with a communication or a particular segment of a communication.
• In some implementations, providing access to the first media analysis result includes: theplatform system300 providing stored communication records via at least one of an API (e.g.,383 ofFIG.3), a browsable user interface, an analytics portal, or any suitable interface. In some implementations, theplatform system300 stores communication records (e.g., call records, messaging records, etc.) of a communication system (e.g., a communication system of theplatform system300, a communication system of the micro-service321, a communication system external to the system300). In some implementations, thesystem300 provides an external system (e.g.,351-355) with the stored communication records via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface, the communication records being communication records of an account of the platform system (e.g., an account managed by the account system330) that is associated with the external system. In some implementations, the external system provides an API request (e.g., via the API383) to receive communication records for an account specified in the API request, and the platform system provides the communication records for the account to the external system via the API of the platform.
• In some implementations, the communication records include information for at least one communication session, and information for each communication session includes at least one of: endpoint information for each communication endpoint of the communication session, an account or sub-account associated with the communication session, endpoint device information (e.g., of adevice356,357 ofFIG.3) for at least one communication endpoint of the communication session, a time period of the communication session, at least one location associated with the communication session, and media analysis results associated with the communication session. In some implementations, media analysis results included in communication records includes the first media analysis result. In some implementations, media analysis results included in communication records includes transcription information for a communication session, language detection information for a communication session, sentiment detection information for a communication session, emotion detection information for a communication session, content classification information for a communication session, context information for a communication session, and speaker detection information for a communication session. In some implementations, media analysis results are stored in the communication records as tags associated with a communication session (or a particular segment of a communication session). In some implementations, transcription information for a communication session (that is included in communication records, e.g., communication records for a platform account) is indexed for searchability.
• In some implementations, a request by an external system (e.g., one of the app servers351-355) for the stored communication records (e.g., provided via at least one of an API, a browsable user interface, an analytics portal, or any suitable interface) (e.g., communication records of a platform system account of the external system) specifies media analysis results properties, and the platform system provides the external system with communication records that match the specified media analysis results properties. In some implementations, the provided communication records for each communication session include media analysis results information that matches the media analysis results properties specified in the request provided by the external system. In some implementations, the platform system (e.g.,300) indexes the media analysis results information of the communication records such that the media analysis results information can be queried. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger. As an example, an entity can provide a request for communication records of communication sessions in which results of a sentiment and emotion detection media analysis indicates anger and results of a transcription media analysis service indicates that overbilling was discussed, and communication sessions in the communication records can be ordered by highest probability of anger to lower probability of anger.
Additional Embodiments
• In some embodiments, the process S510 is similar to S110 ofFIG.2. In some embodiments, the process S520 is similar to S530 ofFIG.2. In some embodiments, the media of the active communication session is collected as described above for S120 ofFIG.2. In some embodiments, the process S530 is similar to S140 ofFIG.2. In some embodiments, the process S510 is similar to S410 ofFIG.4. In some embodiments, the process S520 is similar to S520 ofFIG.4. In some embodiments, the process S530 is similar to S430 ofFIG.4. In some embodiments, the process S550 is similar to S450 ofFIG.4.
• In some embodiments, the first media analysis service is a sentiment and emotion detection service. The active communication session is a communication session between at least a first communication endpoint device (e.g.,356 ofFIG.3) and a second communication endpoint device (e.g.,357 ofFIG.3), the first communication endpoint device and the second communication endpoint device each being external to the platform system (e.g.,300 ofFIG.3). The first media analysis service is constructed to perform media analysis on media of the communication session between at least the first communication endpoint device and the second communication endpoint device. The active communication session is at least one of: a peer-to-peer communication session; and a real-time communication session, and the media is a real-time media stream.
• In some embodiments, the platform system activates a second media analysis service of the plurality of media analysis services. The platform system activates the second media analysis service responsive to a REST API call provided by an external system (e.g.,351-355) of the entity. The second activated media analysis service performs a second media analysis to generate a second media analysis result.
• In some embodiment, the media is real-time media, and the first media analysis service performs real-time media analysis on the real-time media; and the second activated media analysis service performs real-time media analysis on at least one of the real-time media and the first media analysis result.
• In some embodiments, during the active communication session, the platform system (e.g.,300) applies the second media analysis result by transmitting at least one application layer protocol message to a communication service (e.g., of themicro-service321 ofFIG.3) of the platform system, the at least one application layer protocol message providing the second media analysis result.
• In some embodiments, the second media analysis is different from the first media analysis.
• In some embodiments, the media is real-time media, and the first media analysis service performs real-time media analysis on the real-time media; and the second activated media analysis service performs asynchronous media analysis on stored media.
• In some embodiments, activating the first media analysis service includes: the platform system selecting at least the sentiment and emotion detection service as the first media analysis service from the plurality of media analysis services of the platform system, the plurality of media analysis services including the sentiment and emotion detection service and at least one of a context analysis service, and a computer vision service; and the platform system activating each selected media analysis service.
• In some embodiments, the REST API specifies at least the communication session and at least the first activated media analysis service.
• In some embodiments, themethod500 includes at least one of: at the multi-tenant media processing platform system, de-activating the first media analysis service responsive to a de-activation REST API call provided by an external system of the entity, the de-activation REST API call specifying the communication session; and at the multi-tenant media processing platform system, configuring the first media analysis service responsive to a configuration REST API call provided by an external system of the entity, the configuration REST API call specifying the communication session.
• In some embodiment, theplatform system300 trains at least one media analysis service by using the first media analysis result and user feedback provided by an external system (e.g.,351-355) via a feedback API of the platform system. The platform system receives user feedback via the feedback API, the feedback indicating a quality of media analysis provided by the first media analysis service, and the platform system trains the first media analysis service based on the received feedback.
• In some embodiments, the platform system maintains history data that includes media analysis results and corresponding media and user feedback of the first media analysis service, and the platform system trains the first media analysis service by using the history data.
• In some embodiments, the REST API specifies program instructions for a customized media analysis service, and responsive to the REST API request, the platform system performs at least one of: generation of a customized media analysis service according to the received program instructions; and control of a communication service of the platform system to provide a media analysis service according to the received program instructions.
• Method Operation
• In some implementations, the process S510 is performed by theAPI system383. In some implementations, the process S510 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S510 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S510 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S510 is performed by theoperational services399. In some implementations, the process S510 is performed by theaccount system330. In some implementations, the process S510 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S510 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S520 is performed by at least one media analysis micro-service (e.g.,322,323). In some implementations, the process S520 is performed by theAPI system383 and at least one media analysis micro-service. In some implementations, the process S520 is performed by theAPI system383, at least one media analysis micro-service, and the operations services399.
• In some implementations, the process S530 is performed by theAPI system383. In some implementations, the process S530 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S530 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S530 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S530 is performed by theoperational services399. In some implementations, the process S530 is performed by theaccount system330. In some implementations, the process S530 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S530 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S540 is performed by theAPI system383. In some implementations, the process S540 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S540 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S540 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S540 is performed by theoperational services399. In some implementations, the process S540 is performed by theaccount system330. In some implementations, the process S540 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S540 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
• In some implementations, the process S550 is performed by theAPI system383. In some implementations, the process S550 is performed by at least one micro-service (e.g.,321,322,323). In some implementations, the process S550 is performed by theAPI system383 and at least one micro-service. In some implementations, the process S550 is performed by theAPI system383, at least one micro-service, and the operations services399. In some implementations, the process S550 is performed by theoperational services399. In some implementations, the process S550 is performed by theaccount system330. In some implementations, the process S550 is performed by theaccount system330 and thepolicy engine331. In some implementations, the process S550 is performed by two or more of theaccount system330, thepolicy engine331, and theMS interface380.
6. Multi-Tenant MediaProcessing Platform Method600
• As shown inFIG.6, themethod600 is performed at a multi-tenant media processing platform system (e.g., thesystem300 ofFIG.3A) that includes a plurality of media analysis services (e.g., implemented as micro-services322-323 ofFIG.3A).
• In some embodiments, themethod600 is similar to embodiments of themethods400 and500 as described herein.
• The method600 includes: activating a first media analysis service (e.g., of322 ofFIG.3) and a second media analysis service (e.g., of323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session (e.g., a communication session of the communication micro-service321) of an entity (e.g., an entity of one of the systems351-355) in the platform system300 (process S610); the first activated media analysis service performing a first media analysis on the collected media while the communication session is active to generate a first media analysis result (process S620); the second activated media analysis service performing a second media analysis on the collected media while the communication session is active to generate a second media analysis result, the second media analysis being different from the first media analysis (process S630); during the active communication session, applying the first media analysis result and the second media analysis result (process S640); accounting for the use of the first activated media analysis service on behalf of the entity (process S650); accounting for the use of the second activated media analysis service on behalf of the entity, the platform system accounting for the use of the second activated media analysis service independently from the accounting for the use of the first activated media analysis service (process S660); generating billing information for the entity based on the accounting for the use of the first activated media analysis service and the accounting for the use of the second activated media analysis service (process S670).
• The platform system activates each media analysis service responsive to at least one of: a determination by theplatform system300 that at least one property of media of the active communication session that is collected by the platform system maps to at least one setting of entity platform configuration that enables media analysis, the entity platform configuration being platform configuration of the entity in theplatform system300; a communication application instruction (as described herein); a REST API call (as described herein); and a SIP message (as described herein). In some embodiments, the REST API call is provided by an external system of the entity. In some embodiments, the communication application instruction is provided by an external system of the entity. In some embodiments, the SIP message is provided by an external system of the entity.
• The platform system applies the first media analysis result and the second media analysis result by at least one of: changing the active communication session in response to at least one of the first media analysis result and the second media analysis result; and transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by the entity platform configuration, the at least one application layer protocol message providing at least one of the first media analysis result and the second media analysis result.
• The entity platform configuration is provided by at least one external system (e.g.,351-355) via an account management interface (e.g., of theplatform API system383 ofFIG.3) of the mediaprocessing platform system300, the account management interface including a least one of an account portal user interface and an account management API.
• The first media analysis service and the second media analysis service are both active during at least a portion of the active communication session.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a context analysis service, and a computer vision service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a language detection service, an intent detection service, a speaker detection service, a context analysis service, and a computer vision service.
• In some embodiments, the active communication session is a communication session between at least a first communication endpoint device (e.g., theendpoint device356 ofFIG.3) and a second communication endpoint device (e.g., theendpoint device357 ofFIG.3), the first communication endpoint device and the second communication endpoint device each being external to theplatform system300, and each media analysis service is constructed to perform media analysis on media of the communication session between at least the first communication endpoint device and the second communication endpoint device.
• In some embodiments, the active communication session is a peer-to-peer communication session.
• In some embodiments, the active communication session is a real-time communication session.
• In some embodiments, the media is a real-time media stream.
• In some embodiments, applying the first media analysis result and the second media analysis result includes transmitting at least one application layer protocol message to at least one callback URI of a communication service of the platform system (e.g., a callback URI of thecommunication micro-service321, such as, for example, a callback URI corresponding to the communication micro-service API391) of the entity, the at least one application layer protocol message providing at least one of the first media analysis result and the second media analysis result.
• In some implementations, a media analysis micro-service (e.g.,323,323) performs at least one of the processes S610-S670. In some implementations, the communication micro-service (e.g.,321) performs at least one of the processes S610-S670. In some implementations, theplatform API system383 performs at least one of the processes S610-S670. In some implementations, theoperational services399 performs at least one of the processes S610-S670.
7. Multi-Tenant MediaProcessing Platform Method700
• As shown inFIG.7, themethod700 is performed at a multi-tenant media processing platform system (e.g., thesystem300 ofFIG.3A) that includes a plurality of media analysis services (e.g., implemented as micro-services322-323 ofFIG.3A).
• In some embodiments, themethod700 is similar to embodiments of themethods400 and500 as described herein.
• Themethod700 includes: activating at least a first media analysis service (e.g., of themicro-service322 ofFIG.3) and a second media analysis service (e.g., of themicro-service323 ofFIG.3) of the plurality of media analysis services for at least a portion of an active communication session of an entity (e.g., of a system351-355) in the platform system (process S710); the first media analysis service collecting media of the active communication session (process S720); the first media analysis service performing a first media analysis on the collected media while the communication session is active to generate a first media analysis result (process S730); the first media analysis service transmitting an application layer protocol message to a Uniform Resource Identifier (URI) of the second media analysis service, the application layer protocol message providing data relating to the first media analysis result (process S740); the second media analysis service performing a second media analysis on the first media analysis result while the communication session is active to generate a second media analysis result, the second media analysis being different from the first media analysis (process S750); during the active communication session, applying the second media analysis result (process S760). A callback URI parameter of the first media analysis service (e.g., stored as entity platform configuration for the entity) is configured to specify the URI of the second media analysis service. The first media analysis service and the second media analysis service are both active during at least a portion of the active communication session.
• In some embodiments, the application layer protocol message is an API call in accordance with an API of a media analysis micro-service (e.g.,API392,393 ofFIG.3) of the second media analysis service.
• In some embodiments, applying the second media analysis result includes transmitting at least one application layer protocol message to at least one callback URI of an external system (e.g.,351-355) of the entity, the at least one callback URI being specified by a REST API call provided by the external system of the entity, the at least one application layer protocol message providing the second media analysis result.
• In some embodiments, applying the second media analysis result includes transmitting at least one application layer protocol message to at least one callback URI of a communication service of the platform system (e.g., a callback URI of thecommunication micro-service321, such as, for example, a callback URI corresponding to the communication micro-service API391) of the entity, the at least one application layer protocol message providing the second media analysis result.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a transcription service, a language detection service, a content detection service, an intent detection service, a speaker detection service, a context analysis service, a computer vision service, and a natural language processing service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a context analysis service, and a computer vision service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service.
• In some embodiments, the plurality of media analysis services includes at least one of a sentiment and emotion detection service, a language detection service, an intent detection service, a speaker detection service, a context analysis service, and a computer vision service.
• In some embodiments, the active communication session is a communication session between at least a first communication endpoint device (e.g., theendpoint device356 ofFIG.3) and a second communication endpoint device (e.g., theendpoint device357 ofFIG.3), the first communication endpoint device and the second communication endpoint device each being external to theplatform system300, and the first media analysis service is constructed to perform media analysis on media of the communication session between at least the first communication endpoint device and the second communication endpoint device.
• In some embodiments, the active communication session is a peer-to-peer communication session.
• In some embodiments, the active communication session is a real-time communication session.
• In some embodiments, the media is a real-time media stream.
• In some implementations, a media analysis micro-service (e.g.,323,323) performs at least one of the processes S710-S760. In some implementations, the communication micro-service (e.g.,321) performs at least one of the processes S710-S760. In some implementations, theplatform API system383 performs at least one of the processes S710-S760. In some implementations, theoperational services399 performs at least one of the processes S710-S760.
8. System Architecture:Platform System300
• FIG.8 is an architecture diagram of a system (e.g., theplatform system300 ofFIG.3A) according to an implementation in which the system is implemented by a server device. In some implementations, the system is implemented by a plurality of devices. In some implementations, thesystem300 is similar to thesystem100 ofFIG.1.
• Thebus801 interfaces with theprocessors801A-801N, the main memory (e.g., a random access memory (RAM))822, a read only memory (ROM)804, a processor-readable storage medium805, and anetwork device811. In some implementations, thesystem300 includes at least one of a display device and a user input device.
• Theprocessors801A-801N may take many forms, such as ARM processors, X86 processors, and the like.
• In some implementations, the system (e.g.,300) includes at least one of a central processing unit (processor) and a multi-processor unit (MPU).
• Theprocessors801A-801N and themain memory822 form aprocessing unit899. In some embodiments, the processing unit includes one or more processors communicatively coupled to one or more of a RAM, ROM, and machine-readable storage medium; the one or more processors of the processing unit receive instructions stored by the one or more of a RAM, ROM, and machine-readable storage medium via a bus; and the one or more processors execute the received instructions. In some embodiments, the processing unit is an ASIC (Application-Specific Integrated Circuit). In some embodiments, the processing unit is a SoC (System-on-Chip). In some embodiments, the processing unit includes one or more of the operational services, the platform API system, and one or more micro-services.
• Thenetwork adapter device811 provides one or more wired or wireless interfaces for exchanging data and commands between the system (e.g.,300) and other devices, such as an external system (e.g.,351-355). Such wired and wireless interfaces include, for example, a universal serial bus (USB) interface, Bluetooth interface, Wi-Fi interface, Ethernet interface, near field communication (NFC) interface, and the like.
• Machine-executable instructions in software programs (such as an operating system, application programs, and device drivers) are loaded into the memory822 (of the processing unit899) from the processor-readable storage medium805, theROM804 or any other storage location. During execution of these software programs, the respective machine-executable instructions are accessed by at least one ofprocessors801A-801N (of the processing unit899) via thebus801, and then executed by at least one ofprocessors801A-801N. Data used by the software programs are also stored in thememory822, and such data is accessed by at least one ofprocessors801A-801N during execution of the machine-executable instructions of the software programs. The processor-readable storage medium805 is one of (or a combination of two or more of) a hard drive, a flash drive, a DVD, a CD, an optical disk, a floppy disk, a flash storage, a solid state drive, a ROM, an EEPROM, an electronic circuit, a semiconductor memory device, and the like. The processor-readable storage medium805 includes machine-executable instructions (and related data) for anoperating system812,software programs813,device drivers814, and thecommunication platform820 of thesystem300. The machine-executable instructions (and related data) for thecommunication platform820 include machine-executable instructions (and related data) for theoperational services399, theplatform API System383, and the micro-services321-323.
9. System Architecture: External System
• FIG.9 is an architecture diagram of an external system (e.g., an entity external system, such as, for example, one of the entity systems351-353 ofFIG.3A) according to an implementation in which the external system is implemented by a server device. In some implementations, the external system is implemented by a plurality of devices.
• Thebus901 interfaces with theprocessors901A-901N, the main memory (e.g., a random access memory (RAM))922, a read only memory (ROM)904, a processor-readable storage medium905, and anetwork device911. In some implementations, the external system includes a display device and a user input device.
• Theprocessors901A-901N may take many forms, such as ARM processors, X86 processors, and the like.
• In some implementations, the server device includes at least one of a central processing unit (processor) and a multi-processor unit (MPU).
• Theprocessors901A-901N and themain memory922 form aprocessing unit999. In some embodiments, the processing unit includes one or more processors communicatively coupled to one or more of a RAM, ROM, and machine-readable storage medium; the one or more processors of the processing unit receive instructions stored by the one or more of a RAM, ROM, and machine-readable storage medium via a bus; and the one or more processors execute the received instructions. In some embodiments, the processing unit is an ASIC (Application-Specific Integrated Circuit). In some embodiments, the processing unit is a SoC (System-on-Chip).
• Thenetwork adapter device911 provides one or more wired or wireless interfaces for exchanging data and commands between the external system and other devices, such as thesystem300 ofFIG.3A. Such wired and wireless interfaces include, for example, a universal serial bus (USB) interface, Bluetooth interface, Wi-Fi interface, Ethernet interface, near field communication (NFC) interface, and the like.
• Machine-executable instructions in software programs (such as an operating system, application programs, and device drivers) are loaded into the memory922 (of the processing unit999) from the processor-readable storage medium905, theROM904 or any other storage location. During execution of these software programs, the respective machine-executable instructions are accessed by at least one ofprocessors901A-901N (of the processing unit999) via thebus901, and then executed by at least one ofprocessors901A-901N. Data used by the software programs are also stored in thememory922, and such data is accessed by at least one ofprocessors901A-901N during execution of the machine-executable instructions of the software programs. The processor-readable storage medium905 is one of (or a combination of two or more of) a hard drive, a flash drive, a DVD, a CD, an optical disk, a floppy disk, a flash storage, a solid state drive, a ROM, an EEPROM, an electronic circuit, a semiconductor memory device, and the like. The processor-readable storage medium905 includes machine-executable instructions (and related data) for anoperating system912,software programs913,device drivers914, anddata920.
10. Machines
• The systems and methods of the preferred embodiments and variations thereof can be embodied and/or implemented at least in part as a machine configured to receive a computer-readable medium storing computer-readable instructions. The instructions are preferably executed by computer-executable components preferably integrated with the media intelligence platform. The computer-readable medium can be stored on any suitable computer-readable media such as RAMs, ROMs, flash memory, EEPROMs, optical devices (CD or DVD), hard drives, floppy drives, or any suitable device. The computer-executable component is preferably a general or application specific processor, but any suitable dedicated hardware or hardware/firmware combination device can alternatively or additionally execute the instructions.
11. Conclusion
• As a person skilled in the art will recognize from the previous detailed description and from the figures and claims, modifications and changes can be made to the preferred embodiments of the invention without departing from the scope of this invention defined in the following claims.

Claims (20)

What is claimed is:

1. A method comprising:

causing, by one or more processors, an analysis service to perform an analysis of at least a portion of media communicated in a communication session, the analysis service generating a result based on the analysis of at least the portion of the media communicated in the communication session; and

initiating, by the one or more processors, an event based on the result generated by the analysis service.

2. The method ofclaim 1, further comprising:

receiving a command to invoke the analysis service in relation to the communication session; and wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to the received command to invoke the analysis service in relation to the communication session.

3. The method ofclaim 1, wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command to invoke the analysis service in relation to the communication session, the command being received from an application server that corresponds to an account that corresponds to the communication session.

4. The method ofclaim 1, wherein:

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to a device that corresponds to an account that corresponds to the communication session.

5. The method ofclaim 1, wherein;

the initiating of the event based on the result generated by the analysis service includes updating a record that corresponds to an account that corresponds to the communication session.

6. The method ofclaim 1, wherein;

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command received from a device, and

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to the device from which the command is received.

7. The method ofclaim 1, wherein:

the communication session includes an asynchronous sequence of messages; and

the analysis service includes at least one of a sentiment detection service, an emotion detection service, or an intent detection service.

8. A system comprising:

one or more processors; and

a machine-readable medium storing instructions that, when executed by the one or more processors, cause the system to perform operations comprising:

causing an analysis service to perform an analysis of at least a portion of media communicated in a communication session, the analysis service generating a result based on the analysis of at least the portion of the media communicated in the communication session; and

initiating an event based on the result generated by the analysis service.

9. The system ofclaim 8, wherein the operations further comprise:

receiving a command to invoke the analysis service in relation to the communication session; and wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to the received command to invoke the analysis service in relation to the communication session.

10. The system ofclaim 8, wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command to invoke the analysis service in relation to the communication session, the command being received from an application server that corresponds to an account that corresponds to the communication session.

11. The system ofclaim 8, wherein:

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to a device that corresponds to an account that corresponds to the communication session.

12. The system ofclaim 8, wherein:

the initiating of the event based on the result generated by the analysis service includes updating a record that corresponds to an account that corresponds to the communication session.

13. The system ofclaim 8, wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command received from a device; and

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to the device from which the command is received.

14. The system ofclaim 8, wherein:

the communication session includes an asynchronous sequence of messages; and

the analysis service includes at least one of a sentiment detection service, an emotion detection service, or an intent detection service.

15. A non-transitory machine-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

causing an analysis service to perform an analysis of at least a portion of media communicated in a communication session, the analysis service generating a result based on the analysis of at least the portion of the media communicated in the communication session; and

initiating an event based on the result generated by the analysis service.

16. The non-transitory machine-readable medium ofclaim 15, wherein the operations further comprise:

receiving a command to invoke the analysis service in relation to the communication session; and wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to the received command to invoke the analysis service in relation to the communication session.

17. The non-transitory machine-readable medium ofclaim 15, wherein:

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command to invoke the analysis service in relation to the communication session, the command being received from an application server that corresponds to an account that corresponds to the communication session.

18. The non-transitory machine-readable medium ofclaim 15, wherein:

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to a device that corresponds to an account that corresponds to the communication session.

19. The non-transitory machine-readable medium ofclaim 15, wherein;

the initiating of the event based on the result generated by the analysis service includes updating a record that corresponds to an account that corresponds to the communication session.

20. The non-transitory machine-readable medium ofclaim 15, wherein;

the causing of the analysis service to perform the analysis of at least the portion of the media is in response to a command received from a device, and

the initiating of the event based on the result generated by the analysis service includes communicating the result generated by the analysis service to the device from which the command is received.

Images