CN114612193A - Commodity information push method, intermediate service platform, equipment and storage medium - Google Patents
Commodity information push method, intermediate service platform, equipment and storage mediumDownload PDFInfo
- Publication number
- CN114612193A CN114612193ACN202210284302.0ACN202210284302ACN114612193ACN 114612193 ACN114612193 ACN 114612193ACN 202210284302 ACN202210284302 ACN 202210284302ACN 114612193 ACN114612193 ACN 114612193A
- Authority
- CN
- China
- Prior art keywords
- information
- model
- commodity
- recommendation
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Recommending goods or services
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及预测模型技术领域，尤其涉及一种商品信息推送方法、中间服务平台、计算机设备及存储介质。The invention relates to the technical field of prediction models, in particular to a commodity information push method, an intermediate service platform, computer equipment and a storage medium.
背景技术Background technique
随着科学技术的发展，越来越多线下销售方式的业务逐渐开拓线上销售方式，例如通过结合人工智能等技术在线上向用户推荐相应的产品或者商品。With the development of science and technology, more and more offline sales methods are gradually developing online sales methods, such as recommending corresponding products or commodities online to users by combining artificial intelligence and other technologies.
现有技术中，一般是通过设定好的业务推荐规则训练的人工智能算法模型，进而通过该人工智能算法模型进行产品或者商品推荐，但是上述方法存在如下不足：业务推荐规则会随着产品的快速变化而变化，进而需要训练根据新的业务规则不断训练新的模型，如此导致产品或者商品推荐的效率较低，且容易出现推荐不匹配的情况，从而导致产品或者商品推荐的准确率较低。In the prior art, the artificial intelligence algorithm model is generally trained by the set business recommendation rules, and then the artificial intelligence algorithm model is used to recommend products or commodities. Rapid changes and changes require training to continuously train new models according to new business rules, which leads to low product or product recommendation efficiency, and is prone to recommendation mismatches, resulting in low product or product recommendation accuracy. .
发明内容SUMMARY OF THE INVENTION
本发明实施例提供一种商品信息推送方法、中间服务平台、计算机设备及存储介质，以解决现有技术中产品或者商品推荐的准确率较低的问题。Embodiments of the present invention provide a commodity information push method, an intermediate service platform, a computer device and a storage medium, so as to solve the problem of low accuracy of product or commodity recommendation in the prior art.
一种商品信息推送方法，该商品信息推送方法应用于中间服务平台，所述中间服务平台包括主服务器以及至少一个备份服务器，所述方法包括：A method for pushing commodity information, the method for pushing commodity information is applied to an intermediate service platform, wherein the intermediate service platform includes a main server and at least one backup server, and the method includes:
接收预设业务平台发送的商品推荐请求；所述商品推荐请求中包括场景调用信息；Receive a commodity recommendation request sent by a preset business platform; the commodity recommendation request includes scene calling information;
从所述主服务器以及所有所述备份服务器中确定目标服务器；所述目标服务器中搭载有虚拟IP；Determine a target server from the primary server and all the backup servers; the target server carries a virtual IP;
通过所述虚拟IP将所述商品推荐请求发送至所述目标服务器中，并通过所述目标服务器根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息；所述初始推荐信息是所述目标推荐模型根据所述场景调用信息生成的；所述初始推荐信息包括至少一个初始推荐商品；The commodity recommendation request is sent to the target server through the virtual IP, and the target server determines the target recommendation model from the preset model platform according to the scene invocation information and obtains initial recommendation information; the The initial recommendation information is generated by the target recommendation model according to the scene invocation information; the initial recommendation information includes at least one initial recommended commodity;
对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品；Perform commodity filtering on the initial recommended information to filter out target recommended commodities from all the initial recommended commodities;
根据所有所述目标推荐商品生成商品推荐列表，并将所述商品推荐列表发送至所述预设业务平台中。A product recommendation list is generated according to all the target recommended products, and the product recommendation list is sent to the preset business platform.
一种商品信息推送装置，包括：A product information push device, comprising:
推荐指令接收模块，用于接收预设业务平台发送的商品推荐请求；所述商品推荐请求中包括场景调用信息；a recommendation instruction receiving module, configured to receive a commodity recommendation request sent by a preset business platform; the commodity recommendation request includes scene calling information;
服务器选择模块，用于从所述主服务器以及所有所述备份服务器中确定目标服务器；所述目标服务器中搭载有虚拟IP；A server selection module is used to determine a target server from the primary server and all the backup servers; the target server is equipped with a virtual IP;
推荐信息获取模块，用于通过所述虚拟IP将所述商品推荐请求发送至所述目标服务器中，并通过所述目标服务器根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息；所述初始推荐信息是所述目标推荐模型根据所述场景调用信息生成的；所述初始推荐信息包括至少一个初始推荐商品；A recommendation information acquisition module is used to send the commodity recommendation request to the target server through the virtual IP, and to determine the target recommendation model from the preset model platform through the target server according to the scene invocation information and Obtaining initial recommendation information; the initial recommendation information is generated by the target recommendation model according to the scene invocation information; the initial recommendation information includes at least one initial recommended commodity;
推荐信息过滤模块，用于对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品；A recommendation information filtering module, configured to perform commodity filtering on the initial recommended information, so as to filter out target recommended commodities from all the initial recommended commodities;
推荐列表生成模块，用于根据所有所述目标推荐商品生成商品推荐列表，并将所述商品推荐列表发送至所述预设业务平台中。A recommendation list generation module is configured to generate a commodity recommendation list according to all the target recommended commodities, and send the commodity recommendation list to the preset business platform.
一种计算机设备，包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现上述商品信息推送方法。A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the above commodity information push method when the processor executes the computer program.
一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现上述商品信息推送方法。A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, realizes the above-mentioned method for pushing commodity information.
上述商品信息推送方法、中间服务平台、计算机设备及存储介质，该方法通过接收预设业务平台发送的商品推荐请求；所述商品推荐请求中包括场景调用信息；从所述主服务器以及所有所述备份服务器中确定目标服务器；所述目标服务器中搭载有虚拟IP；通过所述虚拟IP将所述商品推荐请求发送至所述目标服务器中，并通过所述目标服务器根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息；所述初始推荐信息是所述目标推荐模型根据所述场景调用信息生成的；所述初始推荐信息包括至少一个初始推荐商品；对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品；根据所有所述目标推荐商品生成商品推荐列表，并将所述商品推荐列表发送至所述预设业务平台中。The above commodity information push method, intermediate service platform, computer equipment and storage medium, the method receives a commodity recommendation request sent by a preset business platform; the commodity recommendation request includes scene invocation information; A target server is determined in the backup server; a virtual IP is carried in the target server; the commodity recommendation request is sent to the target server through the virtual IP, and the target server invokes information according to the scene, automatically A target recommendation model is determined in the preset model platform and initial recommendation information is obtained; the initial recommendation information is generated by the target recommendation model according to the scene calling information; the initial recommendation information includes at least one initial recommended commodity; Perform commodity filtering on the initial recommendation information to screen out target recommended commodities from all the initial recommended commodities; generate a commodity recommendation list according to all the target recommended commodities, and send the commodity recommendation list to the preset business platform .
本发明通过设定一个中间服务平台，且该中间服务平台对外(如预设业务平台)提高统一的服务接口，只需要其它平台按照该服务接口设定的参数形式进行输入即可通过该中间服务平台调用相应的算法模型进行商品推荐，如此提高了商品推荐的效率，也同时降低了重复开发的成本。进一步地，通过中间服务平台完成商品过滤这一操作，即可使得商品推荐融入相应的过滤规则，从而提高了商品推荐的准确性。In the present invention, an intermediate service platform is set, and the intermediate service platform improves a unified service interface externally (such as a preset business platform), and the intermediate service only needs to be input by other platforms according to the parameter form set by the service interface. The platform calls the corresponding algorithm model for product recommendation, which improves the efficiency of product recommendation and reduces the cost of repeated development. Further, by completing the operation of commodity filtering through the intermediate service platform, the commodity recommendation can be integrated into the corresponding filtering rules, thereby improving the accuracy of the commodity recommendation.
附图说明Description of drawings
为了更清楚地说明本发明实施例的技术方案，下面将对本发明实施例的描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the drawings that are used in the description of the embodiments of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention. , for those of ordinary skill in the art, other drawings can also be obtained from these drawings without creative labor.
图1是本发明一实施例中商品信息推送方法的一应用环境示意图；1 is a schematic diagram of an application environment of a method for pushing commodity information in an embodiment of the present invention;
图2是本发明一实施例中商品信息推送方法的一流程图；2 is a flowchart of a method for pushing commodity information in an embodiment of the present invention;
图3是本发明一实施例中中间服务平台的一原理框图；3 is a schematic block diagram of an intermediate service platform in an embodiment of the present invention;
图4是本发明一实施例中计算机设备的一示意图。FIG. 4 is a schematic diagram of a computer device in an embodiment of the present invention.
具体实施方式Detailed ways
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
本发明实施例提供的商品信息推送方法，该商品信息推送方法可应用如图1所示的应用环境中。具体地，该商品信息推送方法应用在商品信息推送系统中，该商品信息推送系统包括如图1所示的预设业务平台、中间服务平台以及预设模型平台，预设业务平台和中间服务平台之间，以及中间服务平台与预设模型平台之间通过网络进行通信，用于解决现有技术中产品或者商品推荐的准确率较低的问题。其中，预设业务平台可以为用户端，是指与服务器相对应,为客户提供本地服务的程序。客户端可安装在但不限于各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备上。中间服务平台可以用独立的服务器或者是多个服务器组成的服务器集群来实现。其中，服务器可以是独立的服务器，也可以是提供云服务、云数据库、云计算、云函数、云存储、网络服务、云通信、中间件服务、域名服务、安全服务、内容分发网络(Content Delivery Network，CDN)、以及大数据和人工智能平台等基础云计算服务的云服务器。The commodity information push method provided by the embodiment of the present invention can be applied in the application environment shown in FIG. 1 . Specifically, the commodity information push method is applied in a commodity information push system, and the commodity information push system includes a preset business platform, an intermediate service platform and a preset model platform as shown in FIG. 1 , a preset business platform and an intermediate service platform Communication between them, and between the intermediate service platform and the preset model platform through the network, is used to solve the problem of low accuracy of product or commodity recommendation in the prior art. Wherein, the preset service platform may be the client, which refers to a program corresponding to the server and providing local services for the client. Clients can be installed on, but not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The intermediate service platform can be implemented by an independent server or a server cluster composed of multiple servers. The server may be an independent server, or may provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, and content delivery networks (Content Delivery Networks). Network, CDN), and cloud servers for basic cloud computing services such as big data and artificial intelligence platforms.
在一实施例中，如图2所示，提供一种商品信息推送方法，以该方法应用在图1中的中间服务平台为例进行说明，包括如下步骤：In one embodiment, as shown in FIG. 2 , a method for pushing commodity information is provided, and the method is applied to the intermediate service platform in FIG. 1 as an example for description, including the following steps:
S10：接收预设业务平台发送的商品推荐请求；所述商品推荐请求中包括场景调用信息。S10: Receive a commodity recommendation request sent by a preset business platform; the commodity recommendation request includes scene calling information.
可以理解地，预设业务平台表征的是业务渠道方的平台，该预设业务平台发送的商品推荐请求可以在用户或者销售人员输入相应的场景调用信息之后自动生成的，也可以是用户或者销售人员将包含场景调用信息的商品推荐请求传输至预设业务平台的，在此不做限定。其中，场景调用信息中包括如场景ID，场景名称(也即模型名称)、推荐个性化信息、是否采用ABTest(AB测试)、版本ID1(模型的其中一个版本)、版本ID2(模型的另一个版本)等信息。It can be understood that the preset business platform represents the platform of the business channel party, and the product recommendation request sent by the preset business platform can be automatically generated after the user or salesperson enters the corresponding scene invocation information, or the user or the salesperson. It is not limited here for personnel to transmit a product recommendation request containing scene call information to the preset business platform. Among them, the scene call information includes, for example, scene ID, scene name (that is, model name), recommended personalized information, whether to use ABTest (AB test), version ID1 (one version of the model), version ID2 (the other version of the model) version), etc.
进一步地，在中间服务平台中对其它平台(如预设业务平台)提供一个统一的对外接口。中间服务平台提供几种输出值的标准形式，如词典、列表等，确定好参数及代表含义之后即可设定中间服务平台的对外接口，该对外接口适用于同类型输入的所有外部接口对模型的调用。避免了每一个模型均需要提供一个不同的对外接口的弊端，实现了预设模型平台中的所有算法模型对外接口规范的统一，支持服务的动态编排组合。Further, in the intermediate service platform, a unified external interface is provided to other platforms (eg, preset service platforms). The intermediate service platform provides several standard forms of output values, such as dictionaries, lists, etc. After the parameters and their meanings are determined, the external interface of the intermediate service platform can be set. The external interface is applicable to all external interfaces of the same type of input to the model. call. It avoids the disadvantage that each model needs to provide a different external interface, realizes the unification of external interface specifications of all algorithm models in the preset model platform, and supports dynamic arrangement and combination of services.
S20：从所述主服务器以及所有所述备份服务器中确定目标服务器；所述目标服务器中搭载有虚拟IP。S20: Determine a target server from the primary server and all the backup servers; the target server carries a virtual IP.
可以理解地，在本发明中，中间服务平台中存在多个服务器，也即该中间服务平台相当于一个服务器组，包括一个主服务器以及多个备份服务器，因此，目标服务器即为主服务器或者任意一个备份服务器。虚拟IP即为公网虚拟IP，该虚拟IP用于转发商品推荐请求；进一步地，虚拟IP预先固定搭载在主服务器上，若主服务器处于服务正常状态，也即该主服务器可以正常执行目标推荐模型推荐等步骤，则该主服务器即可作为目标服务器；若服务器不处于服务正常状态(如服务宕机状态)，则可以将该虚拟IP搭载至任意一个处于服务正常状态的备份服务器中，使得搭载该虚拟IP备份服务器作为目标服务器，并完成目标推荐模型推荐等步骤。如此，即可实现当搭载虚拟IP的服务器(主服务器或者备份服务器)发生故障时，可以快速将虚拟IP转移至其它处于服务正常状态的服务器中，降低了服务器发生故障对服务访问请求响应的影响，从而提高了服务访问请求响应的速率。It can be understood that in the present invention, there are multiple servers in the intermediate service platform, that is, the intermediate service platform is equivalent to a server group, including a main server and multiple backup servers. Therefore, the target server is the main server or any server. a backup server. The virtual IP is the public network virtual IP, and the virtual IP is used to forward the product recommendation request; further, the virtual IP is pre-fixed and mounted on the main server. If the main server is in a normal service state, that is, the main server can normally perform target recommendation. Model recommendation and other steps, the primary server can be used as the target server; if the server is not in a normal state of service (such as a service down state), the virtual IP can be carried to any backup server in a normal state of service, so that Carry the virtual IP backup server as the target server, and complete steps such as target recommendation model recommendation. In this way, when the server (main server or backup server) carrying the virtual IP fails, the virtual IP can be quickly transferred to other servers with normal services, which reduces the impact of server failure on the response to service access requests. , thereby increasing the rate at which service access requests respond.
进一步地，主服务器可以通过发送组播信息，若各备份服务器在一定时间内未接收到主服务器发送的组播信息时，则认为主服务器不处于服务正常状态(如服务宕机状态)；若各备份服务器在一定时间内可以接收到主服务器发送的组播信息时，则认为主服务器处于服务正常状态。在主服务器处于服务正常状态则确定主服务器为目标服务器；在主服务器不是服务正常状态时，即可将虚拟IP转移至其它处于服务正常状态的备份服务器中，使得搭载虚拟IP的备份服务器作为目标服务器。Further, the main server can send multicast information, if each backup server does not receive the multicast information sent by the main server within a certain period of time, it is considered that the main server is not in a normal service state (such as a service down state); if When each backup server can receive the multicast information sent by the primary server within a certain period of time, it is considered that the primary server is in a normal service state. When the main server is in normal service state, the main server is determined as the target server; when the main server is not in normal service state, the virtual IP can be transferred to other backup servers in normal service state, so that the backup server equipped with the virtual IP can be used as the target. server.
S30：通过所述虚拟IP将所述商品推荐请求发送至所述目标服务器中，并通过所述目标服务器根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息；所述初始推荐信息是所述目标推荐模型根据所述场景调用信息生成的；所述初始推荐信息包括至少一个初始推荐商品。S30: Send the commodity recommendation request to the target server through the virtual IP, and determine the target recommendation model from the preset model platform and obtain initial recommendation information through the target server according to the scene invocation information; The initial recommendation information is generated by the target recommendation model according to the scene invocation information; the initial recommendation information includes at least one initial recommended commodity.
可以理解地，预设模型平台中存储着模型接口迭代更新的算法模型，中间服务平台记录这些算法模型的相关信息，以及算法模型和模型接口之间的对应关系。目标推荐模型指的是预设模型平台中的一个或者两个模型，模型的数量根据场景调用信息中的是否采用ABTest这一信息来确定，例如当场景调用信息指示需要采用ABTest时，则表征需要调用同一场景下不同版本的模型。初始推荐信息是指在确定目标推荐模型之后，通过与该目标推荐模型对应的接口调用目标推荐模型，从而实现通过目标推荐模型根据场景调用信息中的推荐个性化信息生成；进一步地，该初始推荐信息中包括至少一个初始推荐商品。It can be understood that the preset model platform stores algorithm models whose model interfaces are iteratively updated, and the intermediate service platform records the relevant information of these algorithm models and the corresponding relationship between the algorithm models and the model interfaces. The target recommendation model refers to one or two models in the preset model platform. The number of models is determined according to whether ABTest is used in the scene call information. For example, when the scene call information indicates that ABTest needs to be used, it indicates the need Call different versions of the model in the same scene. The initial recommendation information refers to calling the target recommendation model through the interface corresponding to the target recommendation model after the target recommendation model is determined, so as to realize the generation of the personalized recommendation information in the scene invocation information through the target recommendation model; further, the initial recommendation The information includes at least one initial recommended product.
具体地，在从所述主服务器以及所有所述备份服务器中确定目标服务器之后，即可通过目标服务器根据该商品推荐请求中的场景调用信息解析出相应的模型需求信息(如上述的场景ID，场景名称等)，从而根据该模型需求信息从预设模型平台中确定相对应的目标推荐模型。进一步地，在确定目标推荐模型之后，可以根据与该目标推荐模型对应的接口调用该目标推荐模型，并将场景调用信息输入至目标推荐模型中，以令该目标推荐模型可以根据场景调用信息输出相应的初始推荐商品，所有初始推荐商品组合形成初始推荐信息。Specifically, after the target server is determined from the main server and all the backup servers, the corresponding model requirement information (such as the above-mentioned scene ID, scene name, etc.), so as to determine the corresponding target recommendation model from the preset model platform according to the model requirement information. Further, after the target recommendation model is determined, the target recommendation model can be called according to the interface corresponding to the target recommendation model, and the scene invocation information is input into the target recommendation model, so that the target recommendation model can output according to the scene invocation information. Corresponding initial recommended products, all initial recommended products are combined to form initial recommended information.
S30：对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品。S30: Perform commodity filtering on the initial recommended information to filter out target recommended commodities from all the initial recommended commodities.
可以理解地，在目标推荐模型输出对应的初始推荐信息之后，用户可能存在自身的购买习惯(例如用户不喜欢某些品牌的商品)，或者是历史推荐过程中用户反馈出该场景下某些商品不适合进行推荐，因此需要对初始推荐信息中的初始推荐商品进行过滤。在本实施例中，通过两种方法对初始推荐信息进行过滤，一种为：根据预设业务平台发送的商品推荐请求中的商品过滤规则信息对初始推荐信息进行商品过滤；其中，商品过滤规则信息为用户或者销售人员自定义的过滤规则，例如该商品过滤规则信息指示商品推荐时需要过滤某一品牌的商品；另一种为根据中间服务平台中所存储的预设场景-规则映射表中与场景调用信息相匹配的场景过滤规则信息对初始推荐信息进行商品过滤；其中，场景过滤规则信息中记录着某些场景下的商品是不允许进行售卖的，因此该类商品推荐给用户是无意义的，因此可以根据该场景过滤规则信息对这部分商品进行过滤。进一步地，上述解释的举例只是规则中的其中一种示例，其余合适的规则也同样适用，在此不作更多的举例解释。Understandably, after the target recommendation model outputs the corresponding initial recommendation information, the user may have his own purchasing habits (for example, the user does not like certain brands of products), or the user feedback some products in the scenario during the historical recommendation process. It is not suitable for recommendation, so it is necessary to filter the initial recommended products in the initial recommendation information. In this embodiment, the initial recommendation information is filtered through two methods, one is: performing commodity filtering on the initial recommended information according to the commodity filtering rule information in the commodity recommendation request sent by the preset business platform; wherein, the commodity filtering rules The information is a filter rule customized by the user or salesperson. For example, the product filter rule information indicates that products of a certain brand need to be filtered when the product is recommended; the other is based on the preset scene-rule mapping table stored in the intermediate service platform. The scene filtering rule information that matches the scene calling information performs commodity filtering on the initial recommendation information; among them, the scene filtering rule information records that commodities in certain scenarios are not allowed to be sold, so it is not recommended to recommend such commodities to users. Therefore, this part of the commodities can be filtered according to the filtering rule information of the scene. Further, the example explained above is only one example of the rules, and other suitable rules are also applicable, and more examples are not explained here.
S40：根据所有所述目标推荐商品生成商品推荐列表，并将所述商品推荐列表发送至所述预设业务平台中。S40: Generate a product recommendation list according to all the target recommended products, and send the product recommendation list to the preset business platform.
具体地，在对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品之后，根据所有目标推荐商品生成商品推荐列表，并将商品推荐列表发送至预设业务平台中，从而实现完整的商品推荐流程。Specifically, after product filtering is performed on the initial recommendation information to screen out target recommended products from all the initial recommended products, a product recommendation list is generated according to all the target recommended products, and the product recommendation list is sent to the preset service platform, so as to realize the complete product recommendation process.
在本实施例中，通过设定一个中间服务平台，且该中间服务平台对外(如预设业务平台)提高统一的服务接口，只需要其它平台按照该服务接口设定的参数形式进行输入即可通过该中间服务平台调用相应的算法模型进行商品推荐，如此提高了商品推荐的效率，也同时降低了重复开发的成本。进一步地，通过中间服务平台完成商品过滤这一操作，即可使得商品推荐融入相应的过滤规则，从而提高了商品推荐的准确性。In this embodiment, by setting an intermediate service platform, and the intermediate service platform improves a unified service interface externally (such as a preset business platform), it is only necessary for other platforms to input according to the parameter form set by the service interface. Through the intermediate service platform, the corresponding algorithm model is invoked for product recommendation, which improves the efficiency of product recommendation and reduces the cost of repeated development. Further, by completing the operation of commodity filtering through the intermediate service platform, the commodity recommendation can be integrated into the corresponding filtering rules, thereby improving the accuracy of the commodity recommendation.
在一实施例中，步骤S20中，也即所述根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息，包括：In one embodiment, in step S20, that is, according to the scene invocation information, the target recommendation model is determined from the preset model platform and the initial recommendation information is obtained, including:
对所述场景调用信息进行解析，得到场景调用表字段；所述场景调用表字段中包括模型调用基本信息以及模型调用版本信息。The scene invocation information is parsed to obtain a scene invocation table field; the scene invocation table field includes model invocation basic information and model invocation version information.
可以理解地，对场景调用信息进行解析可以为对场景调用信息进行实体识别等过程，从而确定出场景调用信息中所包含的模型调用基本信息以及模型调用版本信息。进一步地，在上述说明中指出，场景调用信息中包括如场景ID，场景名称(也即模型名称)、推荐个性化信息、是否采用ABTest(AB测试)、版本ID1(模型的其中一个版本)、版本ID2(模型的另一个版本)等信息，因此，此处指的模型调用基本信息即为场景ID、场景名称、推荐个性化信息以及是否采用ABTest；模型调用版本信息即为版本ID1和版本ID2(若不采用ABTest，则模型调用版本信息中为一个版本ID)。It can be understood that the parsing of the scene invocation information may be a process of performing entity recognition on the scene invocation information, so as to determine the basic model invocation information and model invocation version information contained in the scene invocation information. Further, in the above description, it is pointed out that the scene call information includes such as scene ID, scene name (namely model name), recommended personalized information, whether to adopt ABTest (AB test), version ID1 (one of the versions of the model), Version ID2 (another version of the model) and other information, therefore, the basic model call information referred to here is the scene ID, scene name, recommended personalization information, and whether to use ABTest; model call version information is version ID1 and version ID2 (If ABTest is not used, the model call version information is a version ID).
获取模型维护列表；所述模型维护列表中包括至少一个模型维护三元组；一个所述模型维护三元组中包括模型维护基本信息、模型维护版本信息以及模型接口信息。A model maintenance list is obtained; the model maintenance list includes at least one model maintenance triplet; and one of the model maintenance triples includes model maintenance basic information, model maintenance version information, and model interface information.
可以理解地，在预设模型平台中存储着多个算法模型，且每一个算法模型可能存在不同的版本，因此在中间服务平台的数据库中存储着模型维护列表，该模型维护列表中包括至少一个模型维护三元组，该模型维护三元组用于表征不同的算法模型或者不同的模型版本；其中，一个模型维护三元组中包括模型维护基本信息、模型维护版本信息以及模型接口信息。进一步地，模型维护基本信息可以包括如场景ID，场景名称(也即模型名称)；模型维护版本信息指的是与算法模型相对应的版本信息；模型接口信息是指与每一个不同版本的算法模型或者不同的算法模型对应的接口信息，也即一个算法模型对应于一个调用接口。It is understandable that multiple algorithm models are stored in the preset model platform, and each algorithm model may have different versions. Therefore, a model maintenance list is stored in the database of the intermediate service platform, and the model maintenance list includes at least one A model maintenance triple, which is used to represent different algorithm models or different model versions; a model maintenance triple includes basic model maintenance information, model maintenance version information, and model interface information. Further, the basic information of model maintenance may include, for example, scene ID, scene name (namely, model name); model maintenance version information refers to version information corresponding to the algorithm model; model interface information refers to the algorithm corresponding to each different version Model or interface information corresponding to different algorithm models, that is, an algorithm model corresponds to a calling interface.
根据所述模型调用基本信息、模型调用版本信息、模型维护基本信息以及模型维护版本信息，从所有所述模型维护三元组中确定出目标三元组。According to the model invocation basic information, model invocation version information, model maintenance basic information and model maintenance version information, a target triple is determined from all the model maintenance triples.
具体地，在获取模型维护列表之后，即可将模型调用基本信息与模型维护基本信息进行匹配，以及将模型调用版本信息与模型维护版本信息进行匹配，从而确定出具有与模型调用基本信息相同的模型维护基本信息，以及与模型调用版本信息相同的模型维护版本信息的模型维护三元组，并将该模型维护三元组记录为目标三元组。Specifically, after obtaining the model maintenance list, the basic information of model invocation and the basic information of model maintenance can be matched, and the information of model invocation version and the information of model maintenance version can be matched, so as to determine that the basic information of model invocation is the same as the basic information of model invocation. Model maintenance basic information, and model maintenance triples with the same model maintenance version information as model invocation version information, and record the model maintenance triples as target triples.
根据所述目标三元组中的模型维护基本信息以及模型维护版本信息确定所述目标推荐模型，并根据所述目标三元组中的模型接口信息调用所述目标推荐模型，以令所述目标推荐模型根据所述场景调用信息输出所述初始推荐信息。The target recommendation model is determined according to the basic model maintenance information and model maintenance version information in the target triplet, and the target recommendation model is called according to the model interface information in the target triplet, so that the target The recommendation model outputs the initial recommendation information according to the scene invocation information.
具体地，在根据所述模型调用基本信息、模型调用版本信息、模型维护基本信息以及模型维护版本信息，从所有所述模型维护三元组中确定出目标三元组之后，根据目标三元组中的模型维护基本信息以及模型维护版本信息确定目标推荐模型，并根据目标三元组中所指示与目标推荐模型唯一对应的模型接口信息，调用该目标推荐模型，以令该目标推荐模型根据场景调用信息中的推荐个性化信息进行商品推荐，从而得到初始推荐信息。Specifically, after determining the target triplet from all the model maintenance triples according to the model invocation basic information, model invocation version information, model maintenance basic information and model maintenance version information, according to the target triplet The basic model maintenance information and model maintenance version information in the target recommendation model are determined, and the target recommendation model is called according to the model interface information uniquely corresponding to the target recommendation model indicated in the target triplet, so that the target recommendation model can be based on the scenario. The recommended personalized information in the information is called for product recommendation, so as to obtain the initial recommendation information.
在一实施例中，所述根据所述目标三元组中的模型接口信息调用所述目标推荐模型，以令所述目标推荐模型根据所述场景调用信息输出所述初始推荐信息，包括：In one embodiment, the invoking the target recommendation model according to the model interface information in the target triplet, so that the target recommendation model outputs the initial recommendation information according to the scene invocation information, includes:
根据所述模型调用基本信息以及所述模型调用版本信息，确定与所述场景调用信息对应的接口调用方式。According to the basic model invocation information and the model invocation version information, the interface invocation mode corresponding to the scene invocation information is determined.
可以理解地，在本实施例中，接口调用方式具有两种方式：一种为普通调用方式，也即只需要调用一个算法模型的方式；另一种为分流调用方式，也即需要采用ABTest的方式进行调用，此时在模型调用基本信息中可能会出现两个模型的名称，亦或者模型调用版本信息中可能会出现两个不同的版本ID，如此即可通过模型调用基本信息和模型调用版本信息来确定与场景调用信息对应的接口调用方式。Understandably, in this embodiment, the interface invocation mode has two modes: one is a common invocation mode, that is, only one algorithm model needs to be invoked; In this case, the names of two models may appear in the basic information of model invocation, or two different version IDs may appear in the version information of model invocation, so that the basic information and model invocation version can be called through the model. information to determine the interface invocation method corresponding to the scene invocation information.
在所述接口调用方式为分流调用方式时，通过所述目标服务器根据所述场景调用信息进行推荐人群分类，得到人群分类标签。When the interface invocation mode is a shunting invocation mode, the target server performs recommended crowd classification according to the scene invocation information, and obtains a crowd classification label.
可以理解地，在上述说明中指出场景调用信息中包括推荐个性化信息，在推荐个性化信息中可以包括多人的特征信息；例如在销售人员或者测试人员在进行测试时，即可传输如用户访问页面信息的活跃度程度以及访问板块的相关特征，例如：互动时长、点击率、访问专题等等，进而可以根据场景调用信息中的推荐个性化信息进行推荐人群分类，从而得到人群分类标签。Understandably, it is pointed out in the above description that the scene invocation information includes recommended personalized information, and the recommended personalized information may include the characteristic information of multiple people; The degree of activity of the access page information and the relevant characteristics of the access section, such as: interaction time, click rate, access topics, etc., and then the recommended crowd can be classified according to the recommended personalized information in the scene call information, so as to obtain the crowd classification label.
运用AB测试分发规则策略，通过所述目标服务器根据所述场景调用信息以及所述场景调用信息以及所述人群分类标签，生成AB推荐方案。Using the AB test distribution rule strategy, the target server generates an AB recommendation scheme according to the scene invocation information, the scene invocation information, and the crowd classification label.
可理解地，所述AB测试分发规则策略为可以实时刷新的将多个不同的版本分流给不同的用户的商品推荐策略，人群分组的数量和占比会根据AB测试的结果不断调整，最后确定出需要对所有用户发布的推荐方案，所述函数组装的过程为在与页面信息(可以为需要测试的用户访问的页面信息)对应的主函数代码中定位出需插入与所述人群分类标签匹配的云函数代码插入位置，在插入位置相应的插入各云函数代码，并对插入后的所述主函数进行打包，以及生成AB测试方案的过程，将所述AB测试方案与所述页面信息和所述人群分组标签均关联。Understandably, the AB test distribution rule strategy is a product recommendation strategy that can be refreshed in real time and distributes multiple different versions to different users. The number and proportion of crowd groups will be continuously adjusted according to the results of the AB test, and finally determined. The recommended solution that needs to be released to all users, the process of the function assembly is to locate the corresponding main function code with the page information (the page information that can be accessed by the user to be tested) to be inserted and matched with the crowd classification label. The cloud function code insertion position, insert each cloud function code correspondingly at the insertion position, and package the inserted main function, and generate the process of the AB test plan, compare the AB test plan with the page information and The crowd grouping tags are all associated.
其中，所述主函数代码为与其对应的页面信息升级或者修改所创建的主线的代码，所述云函数代码为开发人员仅需提供简单的函数方法名及代码片段即可上线使用的独立的函数代码，所述云函数代码添加触发其运行的触发器，就可以在所述主函数代码中触发其响应及使用。Wherein, the main function code is the code for updating or modifying the created main line corresponding to the page information, and the cloud function code is an independent function that the developer only needs to provide a simple function method name and code snippet and can be used online The cloud function code adds a trigger that triggers its operation, and then its response and usage can be triggered in the main function code.
根据所述目标三元组中的模型接口信息调用所述目标推荐模型，以令所述目标推荐模型根据所述AB推荐方案输出所述初始推荐信息。The target recommendation model is invoked according to the model interface information in the target triplet, so that the target recommendation model outputs the initial recommendation information according to the AB recommendation scheme.
具体地，由于分流调用方式存在两个不同的模型，或者两个不同版本的模型，因此在分流调用方式中存在两个不同的模型接口信息，进而在根据两个不同的模型接口信息调用两个目标推荐模型之后，即可使得目标推荐模型根据AB推荐方案进行商品推荐，得到两个目标推荐模型输出的初始推荐信息，进而可以根据两个目标推荐模型输出的初始推荐信息对应的用户反馈结果对算法模型进行不断完善，从而不断提高算法模型的推荐准确率。Specifically, since there are two different models or two different versions of the model in the offloading calling method, there are two different model interface information in the offloading calling method, and then the two different model interface information are called according to the two different model interface information. After the target recommendation model, the target recommendation model can be used to recommend products according to the AB recommendation scheme, and the initial recommendation information output by the two target recommendation models can be obtained. The algorithm model is continuously improved to continuously improve the recommendation accuracy of the algorithm model.
在一实施例中，所述根据所述模型调用基本信息、模型调用版本信息、模型维护基本信息以及模型维护版本信息，从所有所述模型维护三元组中确定出目标三元组，包括：In one embodiment, the target triplet is determined from all the model maintenance triples according to the model invocation basic information, model invocation version information, model maintenance basic information and model maintenance version information, including:
将所述模型调用基本信息与所述模型维护基本信息进行匹配，以及将所述模型调用版本信息与所述模型维护版本信息进行匹配。The model invocation basic information is matched with the model maintenance basic information, and the model invocation version information is matched with the model maintenance version information.
将具有与所述模型调用基本信息匹配的模型维护基本信息，且与所述模型调用版本信息匹配的模型维护版本信息的模型维护三元组确定为所述目标三元组。A model maintenance triplet having model maintenance basic information matching the model invocation basic information and model maintenance version information matching the model invocation version information is determined as the target triplet.
可以理解地，在获取模型维护列表之后，即可直接将模型调用基本信息与模型维护列表中的模型维护基本信息进行匹配，以及将模型调用版本信息与模型维护列表中的模型维护版本信息进行匹配，从而确定出具有与模型调用基本信息匹配的模型维护基本信息，且与模型调用版本信息匹配的模型维护版本信息的模型维护三元组确定为目标三元组。Understandably, after obtaining the model maintenance list, you can directly match the basic model invocation information with the basic model maintenance information in the model maintenance list, and match the model invocation version information with the model maintenance version information in the model maintenance list. , so that it is determined that the model maintenance triplet with the model maintenance basic information matching the model invocation basic information and the model maintenance version information matching the model invocation version information is determined as the target triplet.
在一实施例中，所述商品推荐请求中还包括商品过滤规则信息；所述对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品，包括：In one embodiment, the commodity recommendation request further includes commodity filtering rule information; the performing commodity filtering on the initial recommended information to filter out target recommended commodities from all the initial recommended commodities, including:
自预设场景-规则映射表中，获取与所述场景调用信息对应的场景过滤规则信息。Obtain scene filtering rule information corresponding to the scene calling information from the preset scene-rule mapping table.
可以理解地，预设场景-规则映射表可以根据历史推荐记录(如商品推荐之后用户的反馈结果)等提前设定，例如在历史推荐记录中一些商品在相应的场景下是不允许进行售卖的，亦或者一些商品推荐给用户之后没有任何的用户消费记录等，从而构建出预设场景-规则映射表。进一步地，该预设场景-规则映射表中存在至少一组场景-规则的映射关系，一组场景-规则的映射关系表征了在某个场景下相对应的商品过滤规则。Understandably, the preset scene-rule mapping table can be set in advance according to historical recommendation records (such as user feedback results after product recommendation), etc. For example, some products in the historical recommendation records are not allowed to be sold in corresponding scenarios. , or some products are recommended to users without any user consumption records, etc., so as to construct a preset scene-rule mapping table. Further, there is at least one group of scene-rule mapping relationships in the preset scene-rule mapping table, and a group of scene-rule mapping relationships represent corresponding commodity filtering rules in a certain scene.
进一步地，在获取初始推荐信息之后，获取预设场景-规则映射表，并从预设场景-规则映射表中查询到场景调用信息中所包含的模型调用基本信息所对应的场景-规则的映射关系，也即从预设场景-规则映射表中查询包含模型调用基本信息的场景-规则的映射关系，进而提取查询到包含模型调用基本信息的场景-规则的映射关系中的过滤规则得到场景过滤规则信息。Further, after obtaining the initial recommendation information, obtain a preset scene-rule mapping table, and query the scene-rule mapping table corresponding to the model invocation basic information contained in the scene invocation information from the preset scene-rule mapping table. relationship, that is, query the scene-rule mapping relationship containing the basic information of model invocation from the preset scene-rule mapping table, and then extract the filtering rules from the scene-rule mapping relationship containing the basic information of model invocation to obtain the scene filter rule information.
根据所述场景过滤规则信息对所述初始推荐信息进行初次商品过滤，以从所有所述初始推荐商品中筛选出场景推荐商品。Perform primary commodity filtering on the initial recommended information according to the scene filtering rule information, so as to filter out scene recommended commodities from all the initial recommended commodities.
具体地，在获取与所述场景调用信息对应的场景过滤规则信息之后，即可根据场景过滤规则信息对初始推荐信息进行初次商品过滤，例如在场景过滤规则信息中记载了相应的商品名称亦或者与商品对应的链接，因此可以直接从初始推荐信息中查询是否存在相应的商品名称，亦或者通过与商品对应的链接进行查询，从而实现对初始推荐信息的初次商品过滤，进而从所有初始推荐商品中筛选得到场景推荐商品。Specifically, after obtaining the scene filtering rule information corresponding to the scene calling information, the initial recommendation information can be filtered for the first time according to the scene filtering rule information. For example, the scene filtering rule information records the corresponding commodity name or The link corresponding to the product, so you can directly query whether there is a corresponding product name from the initial recommended information, or query through the link corresponding to the product, so as to realize the initial product filtering of the initial recommended information, and then from all the initial recommended products. Filter to get scene recommended products.
根据所述商品过滤规则信息对所述场景推荐商品进行再次商品过滤，以从所有所述场景推荐商品中筛选出所述目标推荐商品。According to the commodity filtering rule information, commodity filtering is performed on the scene recommended commodity again, so as to filter out the target recommended commodity from all the scene recommended commodities.
具体地，在根据所述场景过滤规则信息对所述初始推荐信息进行初次商品过滤，以从所有所述初始推荐商品中筛选出场景推荐商品之后，对商品推荐请求中包括的商品过滤规则信息进行解析，以确定商品过滤规则信息是否为空，例如新用户可能需要更多的商品推荐以供选择，该新用户可能未设定商品过滤规则，进而商品过滤规则信息为空，若为空则无需进行过滤；若不为空，则根据商品过滤规则信息对场景推荐商品进行再次商品过滤，从而从所有场景推荐商品中筛选出目标推荐商品。如此通过两重商品过滤规则，即可完成对推荐商品的过滤，从而提高商品推荐的准确性。Specifically, after the initial product filtering is performed on the initial recommendation information according to the scene filtering rule information, so as to screen out the scene recommended products from all the initial recommended products, the product filtering rule information included in the product recommendation request is performed. Parse to determine whether the product filter rule information is empty. For example, a new user may need more product recommendations for selection. The new user may not have set product filter rules, and then the product filter rule information is empty. If it is empty, no need Filter; if it is not empty, perform product filtering on the recommended products in the scenario again according to the product filtering rule information, so as to filter out the target recommended products from all the recommended products in the scenario. In this way, the filtering of the recommended commodities can be completed through the double commodity filtering rules, thereby improving the accuracy of commodity recommendation.
在一实施例中，所述根据所述商品过滤规则信息对所述场景推荐商品进行再次商品过滤，以从所有所述场景推荐商品中筛选出所述目标推荐商品，包括：In one embodiment, performing commodity filtering on the scenario recommended commodities again according to the commodity filtering rule information, so as to filter out the target recommended commodities from all the scenario recommended commodities, includes:
对所述商品过滤规则信息进行空值检测，以确定所述商品过滤规则信息是否为空。Null value detection is performed on the commodity filtering rule information to determine whether the commodity filtering rule information is empty.
具体地，在根据所述场景过滤规则信息对所述初始推荐信息进行初次商品过滤，以从所有所述初始推荐商品中筛选出场景推荐商品之后，对商品过滤规则信息进行空值检测，也即检测商品过滤规则信息中是否包含相应的规则信息。Specifically, after the initial product filtering is performed on the initial recommendation information according to the scene filtering rule information to screen out the scene recommended products from all the initial recommended products, the product filtering rule information is subjected to null value detection, that is, Check whether the product filtering rule information contains corresponding rule information.
在所述商品过滤规则信息不为空时，根据所述商品过滤规则信息对所述场景推荐商品进行再次商品过滤，得到所述目标推荐商品。When the commodity filtering rule information is not empty, perform commodity filtering on the scenario recommended commodity again according to the commodity filtering rule information to obtain the target recommended commodity.
在所述商品过滤规则信息为空时，将所述场景推荐商品记录为所述目标推荐商品。When the commodity filtering rule information is empty, the scene recommended commodity is recorded as the target recommended commodity.
具体地，在对所述商品过滤规则信息进行空值检测，以确定所述商品过滤规则信息是否为空之后，若商品过滤规则信息不为空，则对商品过滤规则信息进行解析，以确定需要过滤的商品名称或者商品链接等信息，从而根据商品名称或者商品链接对场景推荐商品进行再次商品过滤，如若场景推荐商品为商品过滤规则信息中提及需要过滤的商品，则剔除该场景推荐商品，从而将过滤后得到的场景推荐商品确定为目标推荐商品。Specifically, after performing null value detection on the commodity filtering rule information to determine whether the commodity filtering rule information is empty, if the commodity filtering rule information is not empty, parse the commodity filtering rule information to determine whether the commodity filtering rule information is empty. The filtered product name or product link and other information, so as to perform product filtering on the recommended product in the scenario again according to the product name or product link. If the scenario recommended product is the product that needs to be filtered in the product filtering rule information, the scenario recommended product will be removed. Thus, the scene recommended product obtained after filtering is determined as the target recommended product.
进一步地，若商品过滤规则信息为空，则表征用户暂未设定相应的商品过滤规则，则可以直接将上述得到的场景推荐商品确定为目标推荐商品。Further, if the commodity filtering rule information is empty, it means that the user has not set a corresponding commodity filtering rule for the time being, and the above-obtained scene recommended commodity can be directly determined as the target recommended commodity.
在一实施例中，步骤S20中，也即所述从所述主服务器以及所有所述备份服务器中确定目标服务器，包括：In one embodiment, in step S20, that is, the determining of the target server from the primary server and all the backup servers includes:
检测所述主服务器是否处于服务正常状态；所述主服务器中搭载有所述虚拟IP。It is detected whether the main server is in a normal service state; the virtual IP is carried in the main server.
具体地，主服务器可以通过发送组播信息，若各备份服务器在一定时间内未接收到主服务器发送的组播信息时，则认为主服务器不处于服务正常状态(如服务宕机状态)；若各备份服务器在一定时间内可以接收到主服务器发送的组播信息时，则认为主服务器处于服务正常状态。在主服务器处于服务正常状态则确定主服务器为目标服务器；在主服务器不是服务正常状态时，即可将虚拟IP转移至其它处于服务正常状态的备份服务器中，使得搭载虚拟IP的备份服务器作为目标服务器。Specifically, the main server can send multicast information. If each backup server does not receive the multicast information sent by the main server within a certain period of time, it is considered that the main server is not in a normal service state (such as a service down state); if When each backup server can receive the multicast information sent by the primary server within a certain period of time, it is considered that the primary server is in a normal service state. When the main server is in normal service state, the main server is determined as the target server; when the main server is not in normal service state, the virtual IP can be transferred to other backup servers in normal service state, so that the backup server equipped with the virtual IP can be used as the target. server.
在所述主服务器处于服务正常状态时，将所述主服务器记录为所述目标服务器。When the primary server is in a normal service state, the primary server is recorded as the target server.
具体地，在检测主服务器是否处于服务正常状态之后，若主服务器处于服务正常状态，则该主服务器作为目标服务器执行如目标推荐模型选择等操作。Specifically, after detecting whether the primary server is in a normal service state, if the primary server is in a normal service state, the primary server acts as a target server to perform operations such as target recommendation model selection.
在所述主服务器不处于服务正常状态时，将所述虚拟IP转移至任意一个处于服务正常状态的备份服务器中，并将搭载有所述虚拟IP的备份服务器记录为所述目标服务器。When the primary server is not in a normal service state, the virtual IP is transferred to any backup server in a normal service state, and the backup server carrying the virtual IP is recorded as the target server.
具体地，在检测主服务器是否处于服务正常状态之后，若主服务器不处于服务正常状态，也即如主服务器处于服务宕机状态时，检测各备份服务器是否处于服务正常状态，进而可以将虚拟IP转移至任意一个备份服务器中，并且搭载有虚拟IP的备份服务器即作为目标服务器执行如目标推荐模型选择等操作。Specifically, after detecting whether the main server is in a normal service state, if the main server is not in a normal service state, that is, when the main server is in a service down state, it is detected whether each backup server is in a normal service state, and then the virtual IP Transfer to any backup server, and the backup server with virtual IP is used as the target server to perform operations such as target recommendation model selection.
应理解，上述实施例中各步骤的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。It should be understood that the size of the sequence numbers of the steps in the above embodiments does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
在一实施例中，提供一种中间服务平台，该中间服务平台与上述实施例中商品信息推送方法一一对应。如图3所示，该中间服务平台包括推荐指令接收模块10、服务器选择模块20、推荐信息获取模块30、推荐信息过滤模块40和推荐列表生成模块50。各功能模块详细说明如下：In one embodiment, an intermediate service platform is provided, and the intermediate service platform is in one-to-one correspondence with the commodity information pushing method in the above-mentioned embodiment. As shown in FIG. 3 , the intermediate service platform includes a recommendation
推荐指令接收模块10，用于接收预设业务平台发送的商品推荐请求；所述商品推荐请求中包括场景调用信息；The recommendation
服务器选择模块20，用于从主服务器以及所有备份服务器中确定目标服务器；所述目标服务器中搭载有虚拟IP；The
推荐信息获取模块30，用于通过所述虚拟IP将所述商品推荐请求发送至所述目标服务器中，并通过所述目标服务器根据所述场景调用信息，自预设模型平台中确定目标推荐模型并获取初始推荐信息；所述初始推荐信息是所述目标推荐模型根据所述场景调用信息生成的；所述初始推荐信息包括至少一个初始推荐商品；The recommendation
推荐信息过滤模块40，用于对所述初始推荐信息进行商品过滤，以从所有所述初始推荐商品中筛选出目标推荐商品；A recommendation
推荐列表生成模块50，用于根据所有所述目标推荐商品生成商品推荐列表，并将所述商品推荐列表发送至所述预设业务平台中。The recommendation
优选地，推荐信息获取模块30包括：Preferably, the recommendation
信息解析单元，用于对所述场景调用信息进行解析，得到场景调用表字段；所述场景调用表字段中包括模型调用基本信息以及模型调用版本信息；an information parsing unit, configured to parse the scene invocation information to obtain a scene invocation table field; the scene invocation table field includes model invocation basic information and model invocation version information;
列表获取单元，用于获取模型维护列表；所述模型维护列表中包括至少一个模型维护三元组；一个所述模型维护三元组中包括模型维护基本信息、模型维护版本信息以及模型接口信息；a list acquisition unit, configured to acquire a model maintenance list; the model maintenance list includes at least one model maintenance triplet; one of the model maintenance triples includes model maintenance basic information, model maintenance version information, and model interface information;
信息匹配单元，用于根据所述模型调用基本信息、模型调用版本信息、模型维护基本信息以及模型维护版本信息，从所有所述模型维护三元组中确定出目标三元组；an information matching unit, configured to determine a target triplet from all the model maintenance triples according to the model invocation basic information, model invocation version information, model maintenance basic information and model maintenance version information;
模型调用单元，用于根据所述目标三元组中的模型维护基本信息以及模型维护版本信息确定所述目标推荐模型，并根据所述目标三元组中的模型接口信息调用所述目标推荐模型，以令所述目标推荐模型根据所述场景调用信息输出所述初始推荐信息。A model calling unit, configured to determine the target recommendation model according to the model maintenance basic information and model maintenance version information in the target triplet, and call the target recommendation model according to the model interface information in the target triplet , so that the target recommendation model outputs the initial recommendation information according to the scene invocation information.
优选地，模型调用单元包括：Preferably, the model calling unit includes:
接口调用方式确定子单元，用于根据所述模型调用基本信息以及所述模型调用版本信息，确定与所述场景调用信息对应的接口调用方式；an interface invocation mode determination subunit, configured to determine an interface invocation mode corresponding to the scene invocation information according to the model invocation basic information and the model invocation version information;
人群分类子单元，用于在所述接口调用方式为分流调用方式时，通过所述目标服务器根据所述场景调用信息进行推荐人群分类，得到人群分类标签；a crowd classification subunit, configured to classify the recommended crowds according to the scene invocation information through the target server when the interface invocation mode is the diversion invocation mode, and obtain a crowd classification label;
推荐方案确定子单元，用于运用AB测试分发规则策略，通过所述目标服务器根据所述场景调用信息以及所述场景调用信息以及所述人群分类标签，生成AB推荐方案；The recommendation scheme determination subunit is used to use the AB test distribution rule strategy to generate the AB recommendation scheme through the target server according to the scene invocation information, the scene invocation information and the crowd classification label;
模型调用子单元，用于根据所述目标三元组中的模型接口信息调用所述目标推荐模型，以令所述目标推荐模型根据所述AB推荐方案输出所述初始推荐信息。A model calling subunit, configured to call the target recommendation model according to the model interface information in the target triplet, so that the target recommendation model outputs the initial recommendation information according to the AB recommendation scheme.
优选地，信息匹配单元包括：Preferably, the information matching unit includes:
信息匹配子单元，用于将所述模型调用基本信息与所述模型维护基本信息进行匹配，以及将所述模型调用版本信息与所述模型维护版本信息进行匹配；an information matching subunit, configured to match the basic model invocation information with the basic model maintenance information, and match the model invocation version information with the model maintenance version information;
目标三元组确定子单元，用于将与所述模型调用基本信息匹配的模型维护基本信息，且与所述模型调用版本信息匹配的模型维护版本信息对应的模型维护三元组确定为所述目标三元组。The target triplet determination subunit is used to determine the model maintenance basic information that matches the model invocation basic information, and the model maintenance triad corresponding to the model maintenance version information that matches the model invocation version information as the Target triples.
优选地，推荐信息过滤模块40包括：Preferably, the recommended
信息获取单元，用于自预设场景-规则映射表中，获取与所述场景调用信息对应的场景过滤规则信息；an information obtaining unit, configured to obtain scene filtering rule information corresponding to the scene calling information from a preset scene-rule mapping table;
第一过滤单元，用于根据所述场景过滤规则信息对所述初始推荐信息进行初次商品过滤，以从所有所述初始推荐商品中筛选出场景推荐商品；a first filtering unit, configured to perform primary commodity filtering on the initial recommended information according to the scene filtering rule information, so as to filter out scene recommended commodities from all the initial recommended commodities;
第二过滤单元，用于根据所述商品过滤规则信息对所述场景推荐商品进行再次商品过滤，以从所有所述场景推荐商品中筛选出所述目标推荐商品。The second filtering unit is configured to perform commodity filtering on the scenario recommended commodities again according to the commodity filtering rule information, so as to filter out the target recommended commodities from all the scenario recommended commodities.
优选地，第二过滤单元包括：Preferably, the second filter unit includes:
空值检测单元，用于对所述商品过滤规则信息进行空值检测，以确定所述商品过滤规则信息是否为空；A null value detection unit, configured to perform null value detection on the commodity filtering rule information to determine whether the commodity filtering rule information is empty;
第一目标推荐商品确定单元，用于在所述商品过滤规则信息不为空时，根据所述商品过滤规则信息对所述场景推荐商品进行再次商品过滤，得到所述目标推荐商品；a first target recommended commodity determination unit, configured to perform commodity filtering on the scene recommended commodity again according to the commodity filtering rule information when the commodity filtering rule information is not empty, to obtain the target recommended commodity;
第二目标推荐商品确定单元，用于在所述商品过滤规则信息为空时，将所述场景推荐商品记录为所述目标推荐商品。The second target recommended commodity determination unit is configured to record the scene recommended commodity as the target recommended commodity when the commodity filtering rule information is empty.
优选地，服务器选择模块20包括：Preferably, the
服务器检测单元，用于检测所述主服务器是否处于服务正常状态；所述主服务器中搭载有所述虚拟IP。A server detection unit, configured to detect whether the main server is in a normal service state; the virtual IP is carried in the main server.
第一服务器选择单元，用于在所述主服务器处于服务正常状态时，将所述主服务器记录为所述目标服务器；a first server selection unit, configured to record the primary server as the target server when the primary server is in a normal service state;
第二服务器选择单元，用于在所述主服务器不处于服务正常状态时，将所述虚拟IP转移至任意一个处于服务正常状态的备份服务器中，并将搭载有所述虚拟IP的备份服务器记录为所述目标服务器。A second server selection unit, configured to transfer the virtual IP to any backup server in a normal service state when the primary server is not in a normal service state, and record the backup server carrying the virtual IP for the target server.
关于商品信息推送装置的具体限定可以参见上文中对于商品信息推送方法的限定，在此不再赘述。上述商品信息推送装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。For the specific limitation of the commodity information pushing device, please refer to the limitation on the commodity information pushing method above, which will not be repeated here. All or part of the modules in the above commodity information pushing device can be implemented by software, hardware and combinations thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
在一个实施例中，提供了一种计算机设备，该计算机设备可以是服务器，其内部结构图可以如图4所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口和数据库。其中，该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统、计算机程序和数据库。该内存储器为非易失性存储介质中的操作系统和计算机程序的运行提供环境。该计算机设备的数据库用于存储上述实施例中商品信息推送方法所使用到的数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机程序被处理器执行时以实现一种商品信息推送方法。In one embodiment, a computer device is provided, and the computer device may be a server, and its internal structure diagram may be as shown in FIG. 4 . The computer device includes a processor, memory, a network interface, and a database connected by a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store the data used in the method for pushing commodity information in the above embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for pushing commodity information is realized.
在一个实施例中，提供了一种计算机设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，处理器执行计算机程序时实现上述实施例中的商品信息推送方法。In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the method for pushing commodity information in the foregoing embodiment when the computer program is executed. .
在一个实施例中，提供了一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现上述实施例中商品信息推送方法。In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the method for pushing commodity information in the foregoing embodiment is implemented.
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储于一非易失性计算机可读取存储介质中，该计算机程序在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限，RAM以多种形式可得，诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other medium used in the various embodiments provided in this application may include non-volatile and/or volatile memory. Nonvolatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Road (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，仅以上述各功能单元、模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能单元、模块完成，即将所述装置的内部结构划分成不同的功能单元或模块，以完成以上描述的全部或者部分功能。Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. Module completion, that is, dividing the internal structure of the device into different functional units or modules to complete all or part of the functions described above.
以上所述实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围，均应包含在本发明的保护范围之内。The above-mentioned embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it is still possible to implement the foregoing implementations. The technical solutions described in the examples are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the within the protection scope of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210284302.0ACN114612193B (en) | 2022-03-22 | 2022-03-22 | Commodity information push method, intermediate service platform, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210284302.0ACN114612193B (en) | 2022-03-22 | 2022-03-22 | Commodity information push method, intermediate service platform, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114612193Atrue CN114612193A (en) | 2022-06-10 |
| CN114612193B CN114612193B (en) | 2025-05-27 |
Family
ID=81864510
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210284302.0AActiveCN114612193B (en) | 2022-03-22 | 2022-03-22 | Commodity information push method, intermediate service platform, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114612193B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116319648A (en)* | 2023-03-23 | 2023-06-23 | 携程旅游网络技术（上海）有限公司 | Message push method, system, device and storage medium based on third-party platform |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090163183A1 (en)* | 2007-10-04 | 2009-06-25 | O'donoghue Hugh | Recommendation generation systems, apparatus and methods |
| CN110532479A (en)* | 2019-09-05 | 2019-12-03 | 北京思维造物信息科技股份有限公司 | A kind of information recommendation method, device and equipment |
| CN112837107A (en)* | 2019-11-22 | 2021-05-25 | 上海哔哩哔哩科技有限公司 | Cross-platform commodity recommendation method and device and computer equipment |
| CN113592607A (en)* | 2021-08-12 | 2021-11-02 | 脸萌有限公司 | Product recommendation method and device, storage medium and electronic equipment |
- 2022
- 2022-03-22CNCN202210284302.0Apatent/CN114612193B/enactiveActive
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090163183A1 (en)* | 2007-10-04 | 2009-06-25 | O'donoghue Hugh | Recommendation generation systems, apparatus and methods |
| CN110532479A (en)* | 2019-09-05 | 2019-12-03 | 北京思维造物信息科技股份有限公司 | A kind of information recommendation method, device and equipment |
| CN112837107A (en)* | 2019-11-22 | 2021-05-25 | 上海哔哩哔哩科技有限公司 | Cross-platform commodity recommendation method and device and computer equipment |
| CN113592607A (en)* | 2021-08-12 | 2021-11-02 | 脸萌有限公司 | Product recommendation method and device, storage medium and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 吴俊杰等: "数据智能:趋势与挑战", 《系统工程理论与实践》, vol. 40, no. 08, 3 August 2020 (2020-08-03), pages 2116 - 2149* |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116319648A (en)* | 2023-03-23 | 2023-06-23 | 携程旅游网络技术（上海）有限公司 | Message push method, system, device and storage medium based on third-party platform |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114612193B (en) | 2025-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11847480B2 (en) | System for detecting impairment issues of distributed hosts | |
| US11379438B2 (en) | Database migration between computing platforms using virtual backups | |
| US11853271B2 (en) | Creating an extensible and scalable data mapping and modeling experience | |
| US11461679B2 (en) | Message management using machine learning techniques | |
| US20180004852A1 (en) | Method and system for facilitating terminal identifiers | |
| CN111247550B (en) | Computerized transaction management module for blockchain networks | |
| CN109445758B (en) | Data processing method, device, computer equipment and storage medium | |
| US20040083426A1 (en) | System and method for generating pre-populated forms | |
| CN114327374A (en) | Business process generation method, device and computer equipment | |
| US20200412671A1 (en) | Chat bot operation and management architecture | |
| CN113961600A (en) | A data query method, device, computer equipment and storage medium | |
| US9503351B1 (en) | Deployment feedback for system updates to resources in private networks | |
| US9378230B1 (en) | Ensuring availability of data in a set being uncorrelated over time | |
| CN110428342B (en) | Data restoration method, server, customer service side and storage medium | |
| US20240070448A1 (en) | Sample data generation for a mocking service | |
| CN114612193A (en) | Commodity information push method, intermediate service platform, equipment and storage medium | |
| CN112231377B (en) | Data mapping method, system, device, server and storage medium | |
| CN119536840A (en) | Configuration file update, microservice configuration file management method, device, equipment and medium | |
| CN113360210A (en) | Data reconciliation method and device, computer equipment and storage medium | |
| CN114020733B (en) | Target data determination method, device, computer equipment and storage medium | |
| US20230267430A1 (en) | Data processing method and device, and computer-readable storage medium | |
| CN116126819A (en) | System log processing method, device and medium | |
| CN116069865A (en) | Inventory sharing method, apparatus, computer device and storage medium | |
| CN115686592A (en) | Mini-program maintenance method, system, storage medium, and computer equipment | |
| CN114579398A (en) | Log storage method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |