Detailed Description
In order to make the technical solutions in the present specification better understood by those skilled in the art, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.
It should be noted that, the information and data related to the user in the embodiments of the present disclosure are information and data authorized by the user or fully authorized by the related parties, and the processes of collecting, storing, using, processing, transmitting, providing, disclosing and applying the related data all comply with relevant laws and regulations and standards, take necessary security measures, do not violate the public welcome, and provide corresponding operation entries for the user or the related parties to select authorization or rejection.
It should also be noted that in the embodiments of the present disclosure, some existing solutions in the industry such as software, components, models, etc. may be mentioned, and they should be considered as exemplary, only for illustrating the feasibility of implementing the technical solution of the present disclosure, but not meant to imply that the applicant has or must not use the solution.
Referring to fig. 1, an embodiment of the present disclosure provides a service data processing method. The method is particularly applied to the server side. In particular implementations, the method may include the following:
s101, receiving and responding to a service popularization attention request, and determining candidate client objects needing attention;
S102, acquiring first-class information data published by candidate client objects, second-class information data published by a transaction service platform and third-class information data published by a third party through a plurality of data sources according to a preset business processing rule at preset time intervals;
S103, combining the first type information data, the second type information data and the third type information data to obtain target joint information data;
S104, processing target joint information data by using a preset business requirement mining model to obtain a corresponding target processing result, wherein the preset business requirement mining model is a model obtained based on training of a large language model;
S105, according to the target processing result, determining the candidate client object with the service requirement as a target client object;
S106, according to the target demand content, determining matched candidate services from a preset candidate service set as target services aiming at target client objects;
S107, creating a target task related to the offline popularization of the target business to the target client object, and generating target prompt information corresponding to the target task;
And S108, sending the target prompt information to a corresponding service terminal, wherein the service terminal receives and responds to the target prompt information and promotes the target service to the target client object on line.
The preset business requirement mining model can be specifically understood as an algorithm model which is obtained by training by utilizing a large language model in advance and can automatically analyze and mine the business requirement of the client object based on the input information of multiple data sources of the client object.
The service requirements may include, in particular, direct service requirements, and/or potential service requirements. The above-mentioned direct service requirement can be understood as a service requirement directly expressed by the client object through the information data. The potential business requirements may be specifically understood as business requirements implicitly expressed by the client object through one or more pieces of related information data, and business requirements that are not present in the current time period but are likely to be present in a relatively close future time period.
The above-mentioned target client object may be specifically understood as a client object having a service requirement and currently being suitable for service popularization. The target service may be specifically understood as a product or service that matches the service requirement of the target client object, and the target client object has a high probability of being accepted. Specifically, for example, the target business may be a financial product, a financial service, or the like. The specification is not limited with respect to the specific contents and types of the target service.
Based on the embodiment, the method and the system can effectively assist the service terminal to accurately and automatically mine and discover the service requirement of the target client object and the target service suitable for the target client object, automatically establish corresponding target tasks in time and prompt the service terminal in a related manner, enable the service terminal to efficiently and accurately realize off-line service popularization, obtain a better popularization effect and simultaneously reduce the workload of service personnel.
In some embodiments, referring to fig. 2, the above-mentioned service data processing method may be specifically applied to a server side. Specifically, the server may include a background server applied to a side of a transaction service platform (for example, XX bank) and capable of realizing functions of data transmission, data processing and the like. Specifically, the server may be, for example, an electronic device having data operation, storage function and network interaction function. Or the server may be a software program running in the electronic device that provides support for data processing, storage, and network interactions. In the present embodiment, the number of servers is not particularly limited. The server may be one server, several servers, or a server cluster formed by several servers.
In this embodiment, the service terminal may specifically include a front end applied to a side of a service person of a transaction service platform and capable of implementing functions such as data acquisition and data transmission. Specifically, the service terminal may be, for example, an electronic device such as a desktop computer, a tablet computer, a notebook computer, a smart phone, and the like. Or the service terminal may be a software application capable of running in the electronic device.
In the specific implementation, the business personnel can log in the business popularization and attention configuration interface of the transaction service platform through the business terminal. In the interface, the business personnel can select the icon of the client object which is responsible for docking and needs to be focused on as a candidate client object. Correspondingly, the service terminal can respond to the operation, acquire the object identification of the selected candidate client object through the promotion attention configuration interface, generate a corresponding service promotion attention request, and send the service promotion attention request to the server. Wherein, the service promotion attention request at least carries the object identification of the candidate client object.
Further, the service personnel can select the icon of the responsible service as the candidate service in the service promotion and attention configuration interface. Correspondingly, the service promotion attention request can also carry the service identifier of the candidate service.
When the method is implemented, the server can respond to a service popularization attention request, determine candidate client objects needing attention according to object identifiers of the candidate client objects, further determine a plurality of data sources needing attention for each candidate client object, and acquire a plurality of information data related to the candidate client objects through the plurality of data sources.
In some embodiments, the acquiring, by a plurality of data sources, the first type of information data published by the candidate client object may include the following when implemented:
S1, inquiring and determining websites, public numbers and social account numbers of candidate client objects according to object identifications of the candidate client objects;
and S2, inquiring information data published by the candidate client object in the current time period according to the website, the public number and the social account number of the candidate client object, and taking the information data as first-class information data.
The first type of information data may be specifically understood as information data issued by the candidate client object. Specifically, the first type of information data may be public number articles, publicity short videos, externally issued planning books, state messages and the like.
It should be noted that the first type of information data listed above is only a schematic illustration. In specific implementation, the first type of information data may further include other types or content of information data according to specific application scenarios and processing requirements. The present specification is not limited to this.
In specific implementation, the server can automatically track the update dynamics of a plurality of external media such as websites, public numbers, social account numbers and the like of the candidate client objects, and acquire the total information data published by the first candidate client object in the current time period as the first type information data every preset time period (for example, every 1 month).
Based on the embodiment, through a plurality of data sources, the total information data issued by the candidate client object can be comprehensively obtained, so that the service demand condition of the candidate client object can be accurately and comprehensively mined later.
In some embodiments, the acquiring, by a plurality of data sources, the second type of information data published by the transaction service platform may include the following when implemented:
S1, determining attribute information of a candidate client object according to an object identifier of the candidate client object;
And S2, screening out information data matched with the attribute information of the candidate client object from the information data released in the current time period and the adjacent effective historical time period of the transaction service platform, and taking the information data as second-class information data.
The second type of information data may be specifically understood as information data related to the candidate client object, which is issued by a transaction service platform to which the server belongs. The above-described effective history period may specifically be a last period adjacent to the current period, or a last period adjacent to the current period, a last period, and the like.
The attribute information of the candidate client object may specifically include static attribute information and/or dynamic attribute information of the candidate client object.
In implementation, the server may obtain, as the static attribute information, feature data of the candidate client object, such as an object type, a user tag, a city where the candidate client object is located, and a month income, by querying a user database of the platform according to the object identifier of the candidate client object.
In the implementation, the server may also query the user history service operation record of the platform according to the object identifier of the candidate client object, obtain the service operation record of the candidate client object in the last history time period (for example, the last half year), and then analyze the big data according to the service operation record of the last history time period, so as to determine the characteristic data such as the service item, the popularization mode, the operation habit and the like which are relatively preferred recently by the candidate client object, as the dynamic attribute information.
In the specific implementation, the server firstly determines an information release channel which can be reached by the candidate client object from a plurality of external information release channels of the transaction service platform according to the static attribute information and/or the dynamic attribute information of the candidate client object, and then searches the total information data which are externally released by the transaction service platform through the target information release channels in the current time period (for example, the current month) and the last time period (for example, the last month) adjacent to the current time period every preset time period, and finds out the information data which are matched with the attribute information of the candidate client object as second-class information data.
Based on the embodiment, the information data which is issued by the transaction service platform and related to the candidate client object can be comprehensively obtained through the plurality of data sources, so that the service demand condition of the candidate client object can be accurately and comprehensively mined later.
In some embodiments, referring to fig. 3, the method for obtaining the third type of information data published by the third party through multiple data sources may include the following when implemented:
S1, determining attribute information of a candidate client object and object identification of an associated object of the candidate client object according to object identification of the candidate client object;
s2, determining a third party public platform related to the candidate client object according to the attribute information of the candidate client object, and determining a website, a public number and a social account number of the associated object according to the object identification of the associated object;
S3, screening out first branch information related to the candidate client object from information data released in the current time period of the third party public platform and the adjacent effective historical time period;
and S4, combining the first branch message and the second branch message to obtain corresponding third-class information data.
The third type of message data may be specifically understood as information data related to the candidate client object, which is issued in a third manner except for the candidate client object and the transaction service platform.
The third party may specifically include an associated object (e.g., a friend, a partner, a parent, etc. of the candidate client object) and a third party common platform (e.g., a regulatory agency, a service agency, a news media, etc. related to the candidate client object) that have an association with the candidate client object.
Accordingly, the third type of information data may specifically include a first branch message related to the candidate client object and published by the associated object (e.g., text, video, audio, etc. related to the candidate client object and published by the associated object through a website, public number, social account number), and a second branch message related to the candidate client object and published by the third party public platform (e.g., news, notification, announcement, rewards list, etc. related to the candidate client object and published by the third party public platform).
Before implementation, the server can collect the business interaction data of each client object based on the transaction service platform in advance, and then construct a knowledge graph of the client object relationship based on the transaction service platform according to the attribute information of each client object and the business interaction data.
In the implementation, the server can determine the associated object associated with the candidate client object by querying the client object relationship knowledge graph according to the object identification of the candidate client object. Meanwhile, according to the attribute information of the candidate client object, a third party public platform related to the candidate client object can be screened out from public platforms which are authorized to inquire and access by the transaction service platform. Further, the server may search the current time period, the total information data issued by the related object and the total information data issued by the third party public platform in the previous time period adjacent to the current time period every a preset time period, and screen out information data matched with the attribute information of the candidate client object as the first branch message and the second branch message. And combining the first branch message and the second branch message to obtain third-class information data.
Based on the embodiment, the information data which is issued by the third party and related to the candidate client object can be comprehensively obtained through the plurality of data sources, so that the service demand condition of the candidate client object can be accurately and comprehensively mined later.
In some embodiments, the first type of information data, the second type of information data and the third type of information data may be combined to obtain comprehensive and rich target joint information data.
In the implementation, the third type of information data is considered to be relative to the first type of information data and the second type of information data, on one hand, the data which is not directly issued by the candidate client object is relatively uncertain with the actual correlation of the candidate client object, and on the other hand, the data which is not issued by the transaction service platform is relatively uncertain with the reliability of the information data. Based on the above consideration, in order to more accurately use the information data of different data sources later, when the first type information data, the second type information data and the third type information data are combined, weight labels aiming at the first type information data, the second type information data and the third type information data can be respectively set, corresponding weight coefficients are set in the weight labels, and the first type information data, the second type information data and the third type information data carrying the weight labels are spliced according to a specified rule to obtain corresponding target joint information data. The weight label can be obtained through big data analysis based on historical data, specifically, the weight coefficient of the first type of information data is usually larger than that of the second type of information data, and the weight coefficient of the second type of information data is usually larger than that of the third type of information data.
Further, for the third type of information data, corresponding sub-weight coefficients can be set for the information data provided by the corresponding data sources according to the relationship intimacy degree between the data sources of the information data and the candidate client objects and the credit values of the data sources. Wherein the weight coefficient of the obtained information data is relatively larger based on the data source with higher self credit value when the relation affinity degree with the candidate client object is higher.
In the implementation, after the target joint information data is obtained, the target joint information data can be preprocessed. Specifically, whether the content of the information data in the target joint information data is empty or not or whether the content of the information data is invalid can be detected first, and the information data with empty content and the invalid information data are deleted.
Further, the data type of the information data in the target syndicated information data may be determined, where the data type includes text-type data (e.g., public number articles, news, etc.), video-type data (e.g., promotional short videos, etc.), audio-type data (e.g., voice call segments, etc.), picture-type data (e.g., promotional pictures, etc.), and so forth.
The target joint information data is split into a first information data group (corresponding to text type data), a second information data group (corresponding to video type data), a third information data group (corresponding to audio type data), and a fourth information data group (corresponding to picture type data) according to the data type.
And aiming at the second information data set, utilizing a pre-trained video analysis model to perform video content identification and extraction to obtain a corresponding content text, and combining the corresponding content text to be used as the processed second information data set. And aiming at the third information data set, performing voice recognition and arrangement by utilizing a pre-trained voice recognition model to obtain a corresponding content text, and combining the content text to be used as the processed third information data set. And for the fourth information data set, performing text character recognition extraction by utilizing a pre-trained OCR recognition model to obtain corresponding content text, and combining the content text to obtain processed fourth information data.
And combining the first information data set, the processed second information data set, the processed third information data set and the processed fourth information data set to obtain the preprocessed target joint information data.
In specific implementation, the preprocessed target joint information data can be stored in a corresponding text library for later retrieval and use of a preset business requirement mining model.
In some embodiments, in implementation, the target joint information data (or the preprocessed target joint information data) may be input into a preset service requirement mining model, so as to analyze and mine the service requirement of the candidate client object. Correspondingly, the preset business demand mining model outputs a plurality of possible business demand prediction contents and corresponding probability values as target processing results through analysis and mining based on the input target joint information data.
In the implementation, the server may further combine the target joint information data with the attribute information of the candidate client object, and then input the combined target joint information data and the attribute information of the candidate client object into a preset service requirement mining model, so as to obtain a target processing result with better accuracy.
The preset business requirement mining model specifically comprises the steps of training a large language model by utilizing a large amount of sample data in advance, adjusting model parameters in a targeted manner according to an output result of the model during training, adjusting the sample data according to the output result, and continuing model training on the adjusted model by utilizing the adjusted sample data until the preset business requirement mining model meeting the requirements is obtained.
In some embodiments, the server may detect whether a probability value of at least one service requirement predicted content is greater than a preset probability threshold according to a target processing result, determine that the candidate client object currently has a service requirement when determining that the probability value of at least one service requirement predicted content is greater than the preset probability threshold, and mark the candidate client object as the target client object. Further, the service demand predicted content with the largest probability value can be screened out from the plurality of service demand predicted contents in the target processing result to serve as the target demand content of the target client object.
In contrast, when the probability value of the predicted content without any business requirement is larger than the preset probability threshold value, the candidate client object can be determined to have no business requirement currently, and at the moment, prompt information about that the candidate client object temporarily does not need to perform business promotion can be generated.
In the implementation, when it is determined that the probability value of at least one service demand predicted content is greater than a preset probability threshold, and a plurality of service demand predicted contents with the largest probability value exist in the target processing result, the service demand predicted content with the largest probability value can be determined as the candidate predicted content. And processing the rejected target joint information data by using a preset business requirement mining model to obtain a corresponding processing result which is used as an auxiliary reference result. And screening one candidate predicted content which is relatively matched with the auxiliary reference result from the plurality of candidate predicted contents as target demand content.
In some embodiments, during implementation, candidate services which are matched with the attribute information of the target client object and are not yet transacted by the target client object can be screened from a large number of candidate services according to the attribute information of the target client object and the service transacting record of the target client object, and then the candidate services are combined to construct a preset candidate service set aiming at the target client object.
The server can acquire service introduction texts of candidate services in a preset candidate service set, map each candidate service introduction text and target demand content into corresponding service feature vectors and demand feature vectors based on corresponding mapping rules, calculate matching degrees of target demand content of each candidate service domain by utilizing each service feature vector and the demand feature vector, and screen candidate services with highest matching degrees from a plurality of candidate services according to the matching degrees to serve as target services aiming at target client objects.
In some embodiments, the server may automatically create a target task on promoting the target business offline to the target client object at the transaction service platform and send the target task to the manager for review. After the manager checks and passes, the server can further generate target prompt information corresponding to the target task after determining to execute the target task, and meanwhile, determine a service terminal matched with the target task (for example, the service terminal in charge of the target client object is docked), and then send the target prompt information to the service terminal.
Accordingly, the service terminal can display the target prompt information to the service personnel. Business personnel can visit the target client object online according to the target prompt information, and carry out specific target business promotion on the target client object.
In addition, the manager can send an assignment instruction about the target task to the server when the manager passes the audit, wherein the assignment instruction can carry the terminal identification of the business terminal assigned by the manager. Correspondingly, the server can respond to the assignment instruction and send the target prompt information to the corresponding service terminal.
In some embodiments, after determining, according to the target processing result, the candidate client object with the service requirement as the target client object, when the method is implemented as shown in fig. 4, the method may further include the following:
s1, determining target demand content of a target client object according to a target processing result;
S2, screening key information data from the first type information data, the second type information data and the third type information data according to target demand content;
and S3, processing the key information data by using a preset semantic extraction model to obtain a key information abstract and a key content text of the key information data.
Accordingly, after the target prompt information is sent to the corresponding service terminal, the method may further include:
And touching the key information abstract, the key content text and the link address of the key information data to the service terminal.
In specific implementation, the information data with relatively high association degree with the target required content (for example, the information data is ranked at the front based on the association degree) can be selected from the first type of information data, the second type of information data and the third type of information data to serve as key information data. And processing the key information data by using a preset semantic extraction model to obtain a corresponding key information abstract and a key content text. Further, only the key information summary, the key content text, and the key information data may be uploaded to an information database (e.g., a message square) of the transaction service platform, and the key information summary, the key content text, and the link address of the key information data based on the information database may be obtained.
Correspondingly, the link address can be added into the target prompt information, and then the target prompt information is sent to the service terminal. The service terminal can synchronously display the corresponding link address while displaying the target prompt information.
Before the business personnel prepares to perform offline business promotion to the target client object, the business personnel can enter the information database by clicking the corresponding link address, browse and refer to the key information abstract, the key content text and the specific content of the key information data so as to be capable of preparing corresponding promotion strategies in advance and then perform offline promotion of the target business to the target client object based on the promotion strategies, thereby obtaining relatively better promotion effects.
The server can further construct a user portrait of the target client object according to the attribute information of the target client object, then combine the user portrait of the target client object, the target demand content, the introduction text of the target service, related key information abstract and key content text to automatically generate a matched promotion strategy through a corresponding strategy decision model, and provide the promotion strategy to the service terminal so that the service terminal can use the promotion strategy as a reference to better realize the online service promotion aiming at the target client object.
In some embodiments, after sending the target prompt information to the corresponding service terminal, referring to fig. 5, when the method is implemented, the method may further include the following:
s1, receiving target popularization records of target service popularization to target client objects in an offline manner uploaded by a service terminal, and target popularization results;
S2, verifying the target popularization result according to the target popularization record to obtain a corresponding target verification result;
And S3, according to the target verification result, under the condition that verification is confirmed to pass, carrying out quality evaluation on offline service popularization of the service terminal aiming at the target client object according to the target popularization record and the target popularization result to obtain a corresponding target evaluation result.
In the specific implementation, after the business personnel completes the target business offline popularization aiming at the target client object, the business terminal can be used for uploading the target popularization record of the target business offline popularization to the target client object and the target popularization result.
The target promotion record may be text data or audio data. The target popularization result can be one of success popularization, failure popularization and undetermined result.
In the specific implementation, the server can carry out cross verification on the target popularization result by using the target popularization record and the target popularization result according to the target popularization record, so that verification on the target popularization result is realized.
If the target popularization record and the target popularization result content are matched with each other, mutual verification can be achieved, the received target popularization record and the target popularization result can be judged to be accurate and effective, verification is further confirmed to pass, and subsequent quality evaluation can be conducted.
In contrast, if the target popularization record and the target popularization result content are not matched and even contradict each other, it can be judged that at least one of the received target popularization record and the target popularization result is wrong, and further it is determined that the verification fails. At this time, an error report prompt about retransmission of the target popularization record and the target popularization result may be sent to the service terminal.
In specific implementation, the server can process the target popularization record and the target popularization result according to a preset evaluation rule so as to perform quality evaluation on offline service popularization behaviors of the service terminal aiming at the target client object, and obtain a corresponding target evaluation result. The preset evaluation rule is determined by clustering sample popularization records and sample popularization results of service terminals meeting the requirement specifications under a large number of different service scenes in advance.
Furthermore, the matched rewarding data can be determined according to the target evaluation result, and then the rewarding data is distributed to the service terminal.
In addition, according to the target evaluation result, the service terminal with the service popularization quality problem (or not meeting the preset requirement) is determined, and the quality prompt information is sent for the service terminal so as to guide the service terminal to adjust the service popularization mode and strategy in a targeted manner.
Based on the embodiment, the quality evaluation of the offline service popularization behavior of the service terminal aiming at the target client object can be automatically and accurately realized.
The service terminal can encrypt the target promotion result by using the held identity private key to obtain ciphertext data of the target promotion result, then adds a terminal identifier based on the service terminal to the ciphertext data of the target promotion result to generate an identity credential, and simultaneously encrypts the target promotion record by using the identity private key to obtain ciphertext data of the target promotion record. And uploading the ciphertext data of the target popularization result and the ciphertext data of the target popularization record carrying the identity certificate to a server.
The server can detect whether the target popularization result carries an identity certificate or not, check the identity certificate, determine a matched identity public key according to the identity certificate under the condition that the identity certificate is determined to be true, and decrypt ciphertext data of the target popularization result and ciphertext data of the target popularization record respectively by using the identity public key to obtain the corresponding plaintext form target popularization record and target popularization result.
In some embodiments, after obtaining the corresponding target evaluation result, referring to fig. 6, when the method is implemented, the method may further include the following:
s1, detecting whether a target client object and a target service meet a re-popularization condition or not under the condition that the offline service popularization of the service terminal aiming at the target client object is determined to be not in accordance with a preset requirement according to a target evaluation result and the target popularization result is popularization failure;
S2, under the condition that the target client object and the target service meet the re-promotion conditions, creating a re-promotion task for re-promoting the target service to the lower part of the target client object;
S3, determining an improved popularization strategy by processing the target popularization record, the attribute information of the service terminal and the attribute information of the target client object by utilizing a preset popularization strategy generation model;
And S4, transmitting a re-popularization prompt message to the service terminal, wherein the re-popularization prompt message at least carries an improved popularization strategy.
The number of times that the service terminal performs the behavior operation that does not conform to the preset evaluation rule in the downlink service popularization process of the target client object is greater than a preset number of times threshold (for example, 5 times).
In the implementation, whether the target client object is an important client can be detected according to the attribute information of the target client object, meanwhile, whether the target service is a key service of a transaction service platform is detected according to the related information of the target service, and under the condition that the target client object is an important client and the target service is a key service, the target client object and the target service are determined to meet the re-popularization condition.
Further, a model can be generated by utilizing a pre-trained preset promotion strategy, and an improved re-promotion strategy aiming at the target promotion record can be generated by jointly processing the target promotion record, the attribute information of the service terminal and the attribute information of the target client object. And then the improved re-promotion strategy is sent to the service terminal through the re-promotion prompt information.
Correspondingly, the service terminal can respond to the re-promotion prompt information, pertinently correct and adjust the behavior operation which does not accord with the preset evaluation rule and exists when the target client object is promoted in a downlink mode before according to the improved re-promotion strategy, visit the target client object again and promote the target service in a downlink mode again so as to obtain a relatively better promotion effect.
Based on the embodiment, the method and the device can effectively avoid important customer loss, improve the popularization success rate of key business for important customers, and simultaneously are beneficial to the improvement of the related business capability of the business terminal.
From the above, before implementation, the service data processing method provided in the embodiments of the present disclosure may be trained by using a large language model in advance to obtain a preset service demand mining model capable of automatically analyzing and mining a service demand of a client object based on information data of the client object of the input multiple data sources. The method comprises the steps of obtaining first type information data related to candidate client objects needing to be concerned and second type information data issued by a transaction service platform and third type information data issued by a third party through a plurality of data sources according to preset business processing rules at preset time intervals, combining the information data of the plurality of data sources to obtain target joint information data, processing the target joint information data by utilizing a preset business requirement mining model to obtain a corresponding target processing result, determining the candidate client objects with business requirements as target client objects according to the target processing result, determining target requirement contents of the target client objects, determining matched target businesses from preset candidate business sets according to the target requirement contents, creating target tasks related to the target businesses which are promoted to the target client objects, generating target prompt information corresponding to the target tasks, and sending the target prompt information to corresponding business terminals. Therefore, the method and the system can effectively assist the service terminal to accurately and automatically mine and discover the service requirement of the target client object and the target service suitable for the target client object, automatically establish corresponding target tasks in time and carry out relevant prompt on the service terminal, so that the service terminal can efficiently and accurately realize off-line service popularization and obtain better popularization effect.
The embodiment of the present disclosure provides a server, and is shown in fig. 7. The server includes a network communication port 701, a processor 702, and a memory 703, where the above structures are connected by an internal cable, so that each structure may perform specific data interaction.
The network communication port 701 may be specifically configured to receive a service promotion attention request.
The processor 702 is specifically configured to respond to a service promotion attention request, determine candidate client objects to be focused, acquire first class information data published by the candidate client objects through a plurality of data sources according to a preset service processing rule at preset time intervals, and second class information data and third class information data published by a transaction service platform, combine the first class information data, the second class information data and the third class information data to obtain target joint information data, process the target joint information data by using a preset service requirement mining model to obtain a corresponding target processing result, wherein the preset service requirement mining model is a model trained based on a large language model, determine candidate client objects with service requirements as target client objects according to the target processing result, determine target requirement contents of the target client objects, determine matched candidate services from a preset candidate service set according to the target requirement contents as target services for the target client objects, create a target task about the target service to be promoted to the target client objects, generate target prompt information corresponding to the target task, send the target prompt information to the target client object on line, and respond to the target prompt information to the target client terminal, and respond to the target service prompt information.
The memory 703 may be specifically configured to store a corresponding instruction program, and related data such as target joint information data, target demand content, and the like.
Based on the method, the relevant structural performance of the server can be effectively utilized, the data processing speed of the electronic equipment is improved, and the service data processing is efficiently realized.
In this embodiment, the network communication port 701 may be a virtual port that binds with different communication protocols, so that different data may be sent or received. For example, the network communication port may be a port responsible for performing web data communication, a port responsible for performing FTP data communication, or a port responsible for performing mail data communication. The network communication port may also be an entity's communication interface or a communication chip. For example, it may be a wireless mobile network communication chip such as GSM, CDMA, etc., it may also be a Wifi chip, it may also be a bluetooth chip.
In this embodiment, the processor 702 may be implemented in any suitable manner. For example, a processor may take the form of, for example, a microprocessor or processor, and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, application SPECIFIC INTEGRATED Circuits (ASICs), programmable logic controllers, and embedded microcontrollers, among others. The description is not intended to be limiting.
In this embodiment, the memory 703 may include multiple levels, and in a digital system, the memory may be any memory as long as binary data can be stored, in an integrated circuit, a circuit with a memory function without a physical form, such as a RAM, a FIFO, etc., and in a system, a memory device with a physical form, such as a memory bank, a TF card, etc.
The embodiment of the specification also provides a computer readable storage medium based on the service data processing method, wherein the computer readable storage medium stores computer program instructions, the computer program instructions are implemented when the computer program instructions are executed, the computer readable storage medium receives and responds to service popularization attention requests, determines candidate client objects needing attention, acquires first type information data issued by the candidate client objects and second type information data issued by a transaction service platform and third type information data issued by a third party through a plurality of data sources according to preset service processing rules at preset time intervals, combines the first type information data, the second type information data and the third type information data to obtain target joint information data, processes the target joint information data by utilizing a preset service demand mining model to obtain corresponding target processing results, wherein the preset service demand mining model is a model obtained by training based on a large language model, determines the candidate client objects with service demands as target client objects according to the target processing results, determines target demand contents of the target client objects, and sends the target client objects from preset candidate service objects as target client objects to the corresponding to a target prompt message of the target client, and the target client objects to be matched with the target client objects, and the target information is generated by the target prompt message corresponding to the target client terminal, and the target prompt is generated to the target client terminal.
In the present embodiment, the storage medium includes, but is not limited to, a random access Memory (Random Access Memory, RAM), a Read-Only Memory (ROM), a Cache (Cache), a hard disk (HARD DISK DRIVE, HDD), or a Memory Card (Memory Card). The memory may be used to store computer program instructions. The network communication unit may be an interface for performing network connection communication, which is set in accordance with a standard prescribed by a communication protocol.
In this embodiment, the functions and effects of the program instructions stored in the computer readable storage medium may be explained in comparison with other embodiments, and are not described herein.
The embodiment of the specification also provides a computer program product, which at least comprises a computer program, when the computer program is executed by a processor, the method comprises the following steps of receiving and responding to a service promotion attention request, determining candidate client objects needing attention, obtaining first type information data issued by the candidate client objects through a plurality of data sources in preset time intervals according to preset service processing rules, obtaining second type information data issued by a transaction service platform and third type information data issued by a third party, combining the first type information data, the second type information data and the third type information data to obtain target joint information data, processing the target joint information data by utilizing a preset service requirement mining model to obtain corresponding target processing results, wherein the preset service requirement mining model is a model trained based on a large language model, determining the candidate client objects with service requirements as target client objects according to the target processing results, determining target requirement contents of the target client objects, determining matched candidate services from preset candidate service sets according to target requirement contents, creating target service sets aiming at the target client objects, corresponding to the target client objects, responding to the service requirements, generating target prompt information corresponding to the target client objects, and sending the target joint information to the target client objects, and prompting the target client objects to the online service requirements, and prompting the target client objects.
Referring to fig. 8, the embodiment of the present disclosure further provides a service data processing apparatus, which may specifically include the following structural modules:
The receiving module 801 may be specifically configured to receive and respond to a service promotion attention request, and determine a candidate client object that needs attention;
The obtaining module 802 may specifically be configured to obtain, according to a preset service processing rule, first type information data published by a candidate client object, second type information data published by a transaction service platform, and third type information data published by a third party through a plurality of data sources at intervals of a preset time period;
The combination module 803 may be specifically configured to combine the first type of information data, the second type of information data, and the third type of information data to obtain target joint information data;
The processing module 804 may be specifically configured to process the target joint information data by using a preset service requirement mining model to obtain a corresponding target processing result, where the preset service requirement mining model is a model obtained based on training of a large language model;
The first determining module 805 may be specifically configured to determine, according to the target processing result, a candidate client object having a service requirement as a target client object;
the second determining module 806 may be specifically configured to determine, according to the target demand content, a matched candidate service from a preset candidate service set, as a target service for the target client object;
the creation module 807 is specifically configured to create a target task related to the offline popularization of the target service to the target client object, and generate target prompt information corresponding to the target task;
The sending module 808 may be specifically configured to send the target prompt information to a corresponding service terminal, where the service terminal receives and responds to the target prompt information to perform promotion of the target service to the target client object on line.
In some embodiments, the obtaining module 802 may be configured to obtain, through a plurality of data sources, first-type information data published by a candidate client object, by querying and determining, according to an object identifier of the candidate client object, a website, a public number, and a social account number of the candidate client object, and querying, according to the website, the public number, and the social account number of the candidate client object, information data published by the candidate client object in a current period of time, as the first-type information data.
In some embodiments, the obtaining module 802 may be configured to obtain the second type of information data published by the transaction service platform through a plurality of data sources, where the second type of information data is obtained by determining attribute information of the candidate client object according to the object identifier of the candidate client object, and screening information data matching the attribute information of the candidate client object from the information data published in the current time period and the adjacent effective historical time period of the transaction service platform.
In some embodiments, the obtaining module 802 may be configured to obtain third type information data published by a third party through a plurality of data sources according to determining attribute information of a candidate client object and object identification of an associated object of the candidate client object according to object identification of the candidate client object, determining a third party public platform related to the candidate client object according to the attribute information of the candidate client object, determining a website, a public number, and a social account number of the associated object according to the object identification of the associated object, screening out a first branch message related to the candidate client object from information data published in a current time period of the third party public platform and an adjacent effective historical time period, screening out a second branch message related to the candidate client object from information data published in the current time period of the website, the public number, and the social account number of the associated object, and combining the first branch message and the second branch message to obtain corresponding third type information data.
In some embodiments, after determining the candidate client object with the service requirement as the target client object according to the target processing result, the device can be used for determining the target requirement content of the target client object according to the target processing result, screening key information data from the first type information data, the second type information data and the third type information data according to the target requirement content, and processing the key information data by using a preset semantic extraction model to obtain a key information abstract and a key content text of the key information data.
Correspondingly, after the target prompt information is sent to the corresponding service terminal, the device can be used for touching the key information abstract, the key content text and the link address of the key information data to the service terminal when being implemented.
In some embodiments, after the target prompt information is sent to the corresponding service terminal, the device can be used for receiving a target promotion record of target service promotion to the target client object in an offline manner uploaded by the service terminal and a target promotion result, performing verification on the target promotion result according to the target promotion record to obtain a corresponding target verification result, and performing quality evaluation on the offline service promotion of the service terminal to the target client object according to the target promotion record and the target promotion result under the condition that verification is determined to pass according to the target verification result to obtain a corresponding target evaluation result when verification is determined to pass.
In some embodiments, after obtaining the corresponding target evaluation result, the device may be further configured to, when the device is specifically implemented, determine that offline service popularization of the service terminal for the target client object does not meet a preset requirement according to the target evaluation result, and if the target popularization result is that popularization fails, detect whether the target client object and the target service meet a re-popularization condition, create a re-popularization task related to re-popularizing the target service offline to the target client object if the target client object and the target service meet the re-popularization condition, acquire attribute information of the service terminal and attribute information of the target client object, determine an improved popularization policy by processing the target popularization record, the attribute information of the service terminal and the attribute information of the target client object by using a preset popularization policy generation model, and send a re-popularization prompt message to the service terminal, where the re-popularization prompt message at least carries the improved popularization policy.
It should be noted that, the units, devices, or modules described in the above embodiments may be implemented by a computer chip or entity, or may be implemented by a product having a certain function. For convenience of description, the above devices are described as being functionally divided into various modules, respectively. Of course, when the present description is implemented, the functions of each module may be implemented in the same piece or pieces of software and/or hardware, or a module that implements the same function may be implemented by a plurality of sub-modules or a combination of sub-units, or the like. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
From the above, based on the service data processing device provided by the embodiment of the specification, the service data processing device can effectively assist the service terminal to accurately and automatically mine and discover the service requirement of the target client object and the target service suitable for the target client object, automatically create a corresponding target task in time, and prompt the service terminal in a related manner, so that the service terminal can efficiently and accurately realize off-line service popularization, and a better popularization effect is obtained.
In a specific scenario example, in implementation, a news business machine (e.g., business requirement) analysis and customer interview (e.g., business popularization of customers) of a bank based on a large financial model (e.g., a preset business requirement mining model) can be implemented according to the business data processing method provided in the present specification. The specific implementation process can be referred to as follows.
In this scenario example, consider that a full-flow AI analysis of customer manager (e.g., business person) from business machine identification (e.g., mining business needs), business machine analysis to customer visit, post-visit summary can be implemented in conjunction with a financial big model.
In this scenario example, referring to fig. 9, news (e.g., third-class information data) in a specified direction may be analyzed by using a large model, and according to news content, a financial large model is called to generate a summary of news and business objects (e.g., client objects) in the news, and business information (e.g., business requirement content) is listed, and then by combining with a small model, a similarity is performed to match a client name in a line, and business information such as a client investigation report of the client is queried.
Meanwhile, according to the business object and business information generated in the last step, an administrator can release business visit tasks through a business analysis system in a targeted manner, and a client manager can consult relevant information of a client through the business analysis system to prepare a better marketing strategy.
After the visit is completed, the client manager can also describe the visit process (such as target popularization record) through voice, the system generates visit content through voice recognition, and generates the visit record of this time to drop and analyze through calling the financial big model. The system also periodically performs AI scoring on the client visit content of the client manager to evaluate the client visit quality, so that the manager systematically assigns and manages the client visit tasks, and the efficiency is improved.
In implementation, referring to fig. 9, after possible news business data is obtained, the system will store news in a warehouse at regular time, and perform object analysis on the news by using AI skills for browsing by the user. Meanwhile, the business machine system can be linked with other management application, so that the availability and high efficiency of the business machine visiting flow of the branch business machine are ensured.
For the acquisition of news business, the main data source of news business analysis and visit is business news provided by business departments. After the user maintains the news hot spot, the corresponding latest business news content is stored in the system, and the text content of the article is recorded in the database at regular time for subsequent system call and user browsing. Subsequently, the data source can be expanded according to specific service requirements. For example, the method comprises the steps of periodically acquiring article contents of a specific website, and searching the existing hot content and keywords through a public interface of a common search engine to acquire the latest relevant article contents.
The news business analysis is mainly divided into four parts, namely data maintenance, news squares, projects and enterprise analysis.
1) Data maintenance
The data maintenance page user can maintain business news by himself. Meanwhile, the user can maintain the concerned project and the line state-owned enterprise business clients by himself so as to enable the AI system to conduct data analysis.
2) News square
The user can view news business article content on a news plaza page. The business machine system calls the skills of the large model scheduling platform, and achieves the functions of abstract generation of business machine news, news description object detection, business machine exploration and the like. Compared with the manual browsing of the business opportunity news to acquire the business opportunity, the business opportunity system reduces the reading quantity required by a user to acquire the business opportunity, and the user can quickly acquire the business opportunity content by reading the abstract and the business opportunity exploration information. A user can add and maintain the news description object by himself or herself to view the specific related content of the news description object in news.
The user can click on the news description object, the system can automatically search the clients in the row corresponding to the matching according to the similarity matching of the news description object, and the detailed information of the clients in the guest group system is displayed to provide reference for the user.
Meanwhile, on a news square page, the business machine system is linked with the client visit system and the client condition express delivery system, and a business machine visit manager can initiate the client condition express delivery business machine pushing of the news description object and release visit tasks and distribute the visit tasks to the management and protection client manager of the client by clicking the news description object.
3) Project/enterprise analysis
According to the project/enterprise maintained by the user on the data maintenance page, the system analyzes the page on the project/enterprise to generate a business opportunity event time axis of the project/enterprise concerned by the user, and the user can browse all business opportunity news and related abstracts on the time axis and historical business opportunity exploration information generated by the AI large model.
Meanwhile, the user can carry out AI summarization and analysis on the concerned project/enterprise in the user-defined time dimension, and can check the contents of the AI large model, such as the periodic news event brief of the project/enterprise, the periodic fund requirement of the project/enterprise, the possible deep cooperation field of the project/enterprise and banks, the related advice of the project/enterprise for the client manager and the like in the selected time dimension.
4) Interview task publication
A user can initiate business visit to a news description object through a business machine system, and push business visit tasks to a management and protection client manager of the in-line client corresponding to the object.
After the business machine visit task is completed, a visit client manager can input the contents such as the progress, the result and the like of the visit through voice, the system generates text contents in real time according to the input voice, and generates the visit contents such as the visit purpose, the client requirement, the client operation condition, the client intention, the bank business law and the like of the visit task by utilizing the platform large model. Meanwhile, the system carries out AI scoring according to the visit content of the time so as to be used for the manager to carry out reference and unified management.
Through the scene examples, the business data processing method provided by the specification is verified, business opportunity acquisition can be well achieved, particularly, aiming at the existing clients and potential clients, the search condition accuracy is high, the crawling efficiency is high, the AI large model is utilized to generate industry characteristic analysis and reports, the existing client data and financial products are combined to conduct characteristic analysis on the clients and generate industry characteristic reports, analysis can be conducted through the internal self-adjusting large model, and the characteristic visiting task is combined to meet the platform visiting requirement.
Although the present description provides method operational steps as described in the examples or flowcharts, more or fewer operational steps may be included based on conventional or non-inventive means. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When implemented by an apparatus or client product in practice, the methods illustrated in the embodiments or figures may be performed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even in a distributed data processing environment). The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, it is not excluded that additional identical or equivalent elements may be present in a process, method, article, or apparatus that comprises a described element. The terms first, second, etc. are used to denote a name, but not any particular order.
Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller can be regarded as a hardware component, and means for implementing various functions included therein can also be regarded as a structure within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.
The description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer-readable storage media including memory storage devices.
From the above description of embodiments, it will be apparent to those skilled in the art that the present description may be implemented in software plus a necessary general hardware platform. Based on such understanding, the technical solutions of the present specification may be embodied essentially in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and include several instructions to cause a computer device (which may be a personal computer, a mobile terminal, a server, or a network device, etc.) to perform the methods described in the various embodiments or portions of the embodiments of the present specification.
Various embodiments in this specification are described in a progressive manner, and identical or similar parts are all provided for each embodiment, each embodiment focusing on differences from other embodiments. The specification is operational with numerous general purpose or special purpose computer system environments or configurations. Such as a personal computer, a server computer, a hand-held or portable device, a tablet device, a multiprocessor system, a microprocessor-based system, a set top box, a programmable electronic device, a network PC, a minicomputer, a mainframe computer, a distributed computing environment that includes any of the above systems or devices, and the like.
Although the present specification has been described by way of example, it will be appreciated by those skilled in the art that there are many variations and modifications to the specification without departing from the spirit of the specification, and it is intended that the appended claims encompass such variations and modifications as do not depart from the spirit of the specification.