WO2020048084A1

Movatterモバイル変換

Info

Publication number: WO2020048084A1
Application number: PCT/CN2019/073297
Authority: WO
Inventors: 乐志能
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2018-09-07
Filing date: 2019-01-27
Publication date: 2020-03-12
Anticipated expiration: 2021-03-07
Also published as: CN109492180A

Abstract

The present application discloses a resource recommendation method and apparatus, a computer device, and a computer-readable storage medium, relating to the technical field of computers, ensuring that a user to whom a product provider recommends a resource is definitely interested in the product provider, while also ensuring that the recommended resource is of interest to the user, thereby avoiding antipathy of a user caused by continued recommendation of previous content to the user even after tastes of the user have changed. The method comprises: determining multiple candidate users according to a first matching rule; extracting, according to a second matching rule, at least one user to receive a recommendation among the multiple candidate users; acquiring historical use data of each user among the at least one user; determining multiple application resources of an application, and determining, by using a preset algorithm and on the basis of the multiple application resources and the historical use data, at least one resource to be recommended for each user among the at least one user; and ranking the at least one resource, and recommending a resource to the user.

Description

Translated fromChinese

本申请要求与2018年9月7日提交中国专利局、申请号为2018110453881、申请名称为“资源推荐方法、装置、计算机设备及计算机可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在申请中。This application claims priority from the Chinese patent application filed with the Chinese Patent Office on September 7, 2018, with an application number of 2018110453881, and the application name is "resource recommendation method, device, computer equipment, and computer-readable storage medium." Incorporated by reference.

技术领域Technical field

本申请涉及计算机技术领域，特别是涉及一种资源推荐方法、装置、计算机设备及计算机可读存储介质。The present application relates to the field of computer technology, and in particular, to a resource recommendation method, device, computer device, and computer-readable storage medium.

背景技术Background technique

随着互联网技术的成熟与发展，智能手机、个人电脑等终端的功能越来越多，通过终端人们不仅可以进行通话发短信等，还可以浏览网页并获取网页中的各种资源。目前，互联网规模不断扩大，终端可提供的资源的种类和数量也不断快速增长，有时用户需要花费大量的时间才能找到自己喜欢的资源，在找到自己喜欢的资源之前用户可能需要浏览大量不感兴趣的资源，这个浏览的过程会造成用户的终端资源过载，导致用户不断流失，因此，终端中的应用通常会为用户提供资源推荐服务。With the maturity and development of Internet technology, smart phones, personal computers and other terminals have more and more functions. Through the terminals, people can not only make calls, send text messages, etc., but also browse the web and obtain various resources on the web. At present, the scale of the Internet is constantly expanding, and the types and number of resources that terminals can provide continue to grow rapidly. Sometimes users need to spend a lot of time to find their favorite resources. Before finding their favorite resources, users may need to browse a large number of Resources, this browsing process will cause the user's terminal resources to be overloaded, resulting in continuous loss of users. Therefore, applications in the terminal usually provide users with resource recommendation services.

相关技术中，应用在为用户提供资源推荐服务时，通常根据诸如分类算法、聚类算法、协调过滤、逻辑回归、神经网络等算法对用户的用户画像进行分析，确定用户可能感兴趣的资源进行推荐。In related technologies, when providing resource recommendation services for users, the user's user profile is usually analyzed according to algorithms such as classification algorithms, clustering algorithms, coordinated filtering, logistic regression, neural networks and other algorithms to determine the resources that users may be interested in. recommend.

在实现本申请的过程中，发明人发现相关技术至少存在以下问题：目前，应用在对用户画像进行分析时，仅采用某一种算法进行分析，且仅对用户的用户画像进行分析，导致对用户的分析较为片面，推荐给用户的资源不够准确，浪费大量的推荐资源。In the process of implementing this application, the inventor found that the related technology has at least the following problems: At present, when analyzing user portraits, only one algorithm is used for analysis, and only the user portraits of users are analyzed. The user's analysis is one-sided, the resources recommended to the user are not accurate enough, and a lot of recommendation resources are wasted.

发明内容Summary of the Invention

有鉴于此，本申请提供了一种资源推荐方法、装置、计算机设备及计算机可读存储介质，主要目的在于解决目前的对用户的分析较为片面，推荐给用户的资源不够准确，浪费大量的推荐资源的问题。In view of this, this application provides a resource recommendation method, device, computer device, and computer-readable storage medium. The main purpose is to solve the current one-sided analysis of users, the resources recommended to users are not accurate enough, and a large number of recommendations are wasted. Problems with resources.

依据本申请第一方面，提供了一种资源推荐方法，该方法包括：根据第一匹配规则，确定多个候选用户，所述第一匹配规则至少包括目标年龄、目标地区和目标性别；根据第二匹配规则，在所述多个候选用户中，提取至少一个待推荐用户，所述第二匹配规则为预设数目或预设忠诚度；获取所述至少一个待推荐用户中每个待推荐用户的历史使用数据，所述历史使用数据至少包括搜索数据、浏览数据以及交易数据；确定应用的多个应用资源，采用预设算法，基于所述多个应用资源和所述历史使用数据，为所述至少一个待推荐用户中每个待推荐用户确定至少一个待推荐资源；对所述至少一个待推荐资源进行评级，并对所述推荐用户进行资源推荐。According to a first aspect of the present application, a resource recommendation method is provided. The method includes: determining a plurality of candidate users according to a first matching rule, where the first matching rule includes at least a target age, a target region, and a target gender; Two matching rules, extracting at least one user to be recommended from among the plurality of candidate users, the second matching rule being a preset number or preset loyalty; obtaining each user to be recommended from the at least one user to be recommended Historical usage data including at least search data, browsing data, and transaction data; determining multiple application resources for an application, using a preset algorithm, based on the multiple application resources and the historical usage data, for all Each of the at least one to-be-recommended user is described to determine at least one to-be-recommended resource; the at least one to-be-recommended resource is rated, and the recommended user is resource-recommended.

在另一个实施例中，所述根据第一匹配规则，确定多个候选用户，包括：获取至少一个用户的至少一个注册信息，将所述至少一个注册信息与所述第一匹配规则进行比对，所述注册信息至少包括用户年龄、用户地区和用户性别；确定与所述第一匹配规则匹配的多个注册信息，将所述多个注册信息对应的多个用户作为所述多个候选用户。In another embodiment, the determining a plurality of candidate users according to the first matching rule includes: obtaining at least one registration information of at least one user, and comparing the at least one registration information with the first matching rule. The registration information includes at least a user ’s age, a user ’s region, and a user ’s gender; determining a plurality of registration information matching the first matching rule, and using the plurality of users corresponding to the plurality of registration information as the plurality of candidate users .

在另一个实施例中，所述根据第二匹配规则，在所述多个候选用户中，提取至少一个待推荐用户，包括：如果所述第二匹配规则为所述预设数目，则在所述多个候选用户中选取预设数目的候选用户作为所述至少一个待推荐用户；或，如果所述第二匹配规则为所述预设忠诚度，则确定所述多个候选用户的多个用户忠诚度，将用户忠诚度与所述预设忠诚度匹配的候选用户作为所述至少一个待推荐用户。In another embodiment, the extracting at least one user to be recommended among the plurality of candidate users according to a second matching rule includes: if the second matching rule is the preset number, Selecting a preset number of candidate users from the plurality of candidate users as the at least one user to be recommended; or, if the second matching rule is the preset loyalty, determining a plurality of the plurality of candidate users User loyalty, and the candidate user whose user loyalty matches the preset loyalty is used as the at least one user to be recommended.

在另一个实施例中，所述确定应用的多个应用资源，采用预设算法，基于所述多个应用资源和所述历史使用数据，为所述至少一个待推荐用户中每个待推荐用户确定至少一个待推荐资源，包括：对所述应用提供的项目数据及产品数据进行统计，确定所述多个应用资源；对于所述至少一个待推荐用户中的每个待推荐用户，采用所述预设算法，计算所述多个应用资源与所述待推荐用户历史使用数据的多个相似度；在所述多个应用资源中提取相似度大于预设相似度的至少一个应用资源作为所述至少一个待推荐资源。In another embodiment, the plurality of application resources of the application are determined using a preset algorithm, based on the plurality of application resources and the historical usage data, for each of the at least one user to be recommended. Determining at least one resource to be recommended includes: performing statistics on project data and product data provided by the application to determine the plurality of application resources; and using each of the at least one user to be recommended, A preset algorithm that calculates a plurality of similarities between the plurality of application resources and the historical usage data of the user to be recommended; and extracts, from the plurality of application resources, at least one application resource with a similarity greater than a preset similarity as the At least one resource to be recommended.

在另一个实施例中，所述对于所述至少一个待推荐用户中的每个待推荐用户，采用所述预设算法，计算所述多个应用资源与所述待推荐用户历史使用数据的多个相似度，包括：对于所述多个应用资源中的任一应用资源，采用所述预设算法，计算所述搜索数据在所述应用资源的内容中所占的第一比例，计算所述浏览数据在所述应用资源的内容中所占的第二比例，计算所述交易数据在所述应用资源的内容中所占的第三比例；分别确定所述搜索数据的第一权重、所述浏览数据的第二权重以及所述交易数据的第三权重；计算所述第一比例与所述第一权重的第一乘积，计算所述第二比例与所述第二权重的第二乘积，计算所述第三比例与所述第三权重的第三乘积；获取所述第一乘积、所述第二乘积以及所述第三乘积的和值，将所述和值作为所述应用资源与所述历史使用数据的相似度。In another embodiment, for each of the at least one user to be recommended, the preset algorithm is used to calculate a multiplicity of the plurality of application resources and the historical usage data of the user to be recommended. The similarity includes: for any one of the plurality of application resources, using the preset algorithm to calculate a first proportion of the search data in the content of the application resources, and calculating the The second proportion of browsing data in the content of the application resource, and the third proportion of the transaction data in the content of the application resource; calculating the first weight of the search data, the A second weight of browsing data and a third weight of the transaction data; calculating a first product of the first ratio and the first weight, calculating a second product of the second ratio and the second weight, Calculate a third product of the third ratio and the third weight; obtain a sum of the first product, the second product, and the third product, and use the sum as the application resource and Said History using similarity data.

在另一个实施例中，所述对所述至少一个待推荐资源进行评级，并对所述推荐用户进行资源推荐，包括：对所述至少一个待推荐资源进行分类，为所述至少一个待推荐资源评级，生成资源优先级列表；将所述资源优先级列表中资源等级大于预设等级的待推荐资源推荐给所述推荐用户。In another embodiment, the rating of the at least one resource to be recommended, and the resource recommendation of the recommended user includes: classifying the at least one resource to be recommended to be the at least one resource to be recommended. Resource rating, generating a resource priority list; recommending to-be-recommended resources a resource level in the resource priority list that is greater than a preset level.

在另一个实施例中，所述对所述至少一个待推荐资源进行分类，为所述至少一个待推荐资源评级，生成资源优先级列表，包括：获取评级标准，所述评级标准包括多个资源等级与相似度范围之间的对应关系；对于所述至少一个待推荐资源中的每个待推荐资源，获取所述待推荐资源的相似度，对所述相似度进行分类，确定所述相似度所属的目标相似度范围；确定所述目标相似度范围指示的资源等级作为所述待推荐资源的资源等级；将所述待推荐资源的资源标识与所述资源等级对应存储，生成所述资源优先级列表。In another embodiment, the classifying the at least one resource to be recommended and generating a resource priority list for the at least one resource to be recommended includes obtaining a rating standard, where the rating standard includes multiple resources The correspondence between the level and the similarity range; for each of the at least one resource to be recommended, the similarity of the resource to be recommended is obtained, the similarity is classified, and the similarity is determined Belonging target similarity range; determining the resource level indicated by the target similarity range as the resource level of the resource to be recommended; storing the resource identifier of the resource to be recommended corresponding to the resource level, and generating the resource priority Level list.

依据本申请第二方面，提供了一种资源推荐装置，该装置包括：According to a second aspect of the present application, a resource recommendation device is provided. The device includes:

第一确定模块，用于根据第一匹配规则，确定多个候选用户，所述第一匹配规则至少包括目标年龄、目标地区和目标性别；提取模块，用于根据第二匹配规则，在所述多个候选用户中，提取至少一个待推荐用户，所述第二匹配规则为预设数目或预设忠诚度；获取模块，用于获取所述至少一个待推荐用户中每个待推荐用户的历史使用数据，所述历史使用数据至少包括搜索数据、浏览数据以及交易数据；第二确定模块，用于确定应用的多个应用资源，采用预设算法，基于所述多个应用资源和所述历史使用数据，为所述至少一个待推荐用户中每个待推荐用户确定至少一个待推荐资源；推荐模块，用于对所述至少一个待推荐资源进行评级，并对所述推荐用户进行资源推荐。A first determining module, configured to determine a plurality of candidate users according to a first matching rule, the first matching rule including at least a target age, a target region, and a target gender; an extraction module, configured to, according to a second matching rule, Among the plurality of candidate users, at least one user to be recommended is extracted, and the second matching rule is a preset number or preset loyalty; and an acquisition module is configured to obtain a history of each of the at least one user to be recommended among the users to be recommended. Usage data, the historical usage data includes at least search data, browsing data, and transaction data; a second determination module, configured to determine multiple application resources of an application, using a preset algorithm based on the multiple application resources and the history Use data to determine at least one resource to be recommended for each of the at least one user to be recommended; a recommendation module configured to rate the at least one resource to be recommended and perform resource recommendation for the recommended user.

在另一个实施例中，所述第一确定模块，包括：In another embodiment, the first determining module includes:

比对子模块，用于获取至少一个用户的至少一个注册信息，将所述至少一个注册信息与所述第一匹配规则进行比对，所述注册信息至少包括用户年龄、用户地区和用户性别；确定子模块，用于确定与所述第一匹配规则匹配的多个注册信息，将所述多个注册信息对应的多个用户作为所述多个候选用户。A comparison submodule, configured to obtain at least one registration information of at least one user, and compare the at least one registration information with the first matching rule, where the registration information includes at least a user age, a user region, and a user gender; A determining submodule, configured to determine a plurality of registration information matching the first matching rule, and use a plurality of users corresponding to the plurality of registration information as the plurality of candidate users.

在另一个实施例中，所述提取模块，用于如果所述第二匹配规则为所述预设数目，则在所述多个候选用户中选取预设数目的候选用户作为所述至少一个待推荐用户；或，如果所述第二匹配规则为所述预设忠诚度，则确定所述多个候选用户的多个用户忠诚度，将用户忠诚度与所述预设忠诚度匹配的候选用户作为所述至少一个待推荐用户。In another embodiment, the extraction module is configured to, if the second matching rule is the preset number, select a preset number of candidate users as the at least one candidate among the plurality of candidate users. Recommend a user; or, if the second matching rule is the preset loyalty, determine a plurality of user loyalty of the plurality of candidate users, and match the user loyalty with the preset loyalty candidate user As the at least one user to be recommended.

在另一个实施例中，所述第二确定模块，包括：In another embodiment, the second determining module includes:

获取子模块，用于对所述应用提供的项目数据及产品数据进行统计，确定所述多个应用资源；计算子模块，用于对于所述至少一个待推荐用户中的每个待推荐用户，采用所述预设算法，计算所述多个应用资源与所述待推荐用户历史使用数据的多个相似度；确定子模块，用于在所述多个应用资源中提取相似度大于预设相似度的至少一个应用资源作为所述至少一个待推荐资源。An acquisition submodule, configured to perform statistics on project data and product data provided by the application, and determine the multiple application resources; and a calculation submodule, configured to, for each to-be-recommended user of the at least one to-be-recommended user, Adopting the preset algorithm to calculate a plurality of similarities between the plurality of application resources and the historical usage data of the user to be recommended; and determining a sub-module for extracting a similarity greater than a preset similarity among the plurality of application resources The at least one application resource of the degree is used as the at least one resource to be recommended.

在另一个实施例中，所述计算子模块，用于对于所述多个应用资源中的任一应用资源，采用所述预设算法，计算所述搜索数据在所述应用资源的内容中所占的第一比例，计算所述浏览数据在所述应用资源的内容中所占的第二比例，计算所述交易数据在所述应用资源的内容中所占的第三比例；分别确定所述搜索数据的第一权重、所述浏览数据的第二权重以及所述交易数据的第三权重；计算所述第一比例与所述第一权重的第一乘积，计算所述第二比例与所述第二权重的第二乘积，计算所述第三比例与所述第三权重的第三乘积；获取所述第一乘积、所述第二乘积以及所述第三乘积的和值，将所述和值作为所述应用资源与所述历史使用数据的相似度。In another embodiment, the calculation sub-module is configured to use the preset algorithm for any one of the plurality of application resources to calculate the search data in the content of the application resource. A first proportion of the calculation, calculating a second proportion of the browsing data in the content of the application resource, and calculating a third proportion of the transaction data in the content of the application resource; A first weight of the search data, a second weight of the browsing data, and a third weight of the transaction data; calculating a first product of the first ratio and the first weight, calculating the second ratio and the The second product of the second weight, calculating a third product of the third ratio and the third weight; obtaining a sum of the first product, the second product, and the third product, The sum value is used as the similarity between the application resource and the historical usage data.

在另一个实施例中，所述推荐模块，包括：In another embodiment, the recommendation module includes:

生成子模块，用于对所述至少一个待推荐资源进行分类，为所述至少一个待推荐资源评级，生成资源优先级列表；推荐子模块，用于将所述资源优先级列表中资源等级大于预设等级的待推荐资源推荐给所述推荐用户。A generation sub-module for classifying the at least one resource to be recommended, generating a resource priority list for the at least one resource to be recommended; a recommendation sub-module for classifying a resource level in the resource priority list greater than A resource to be recommended at a preset level is recommended to the recommended user.

在另一个实施例中，所述生成子模块，用于获取评级标准，所述评级标准包括多个资源等级与相似度范围之间的对应关系；对于所述至少一个待推荐资源中的每个待推荐资源，获取所述待推荐资源的相似度，对所述相似度进行分类，确定所述相似度所属的目标相似度范围；确定所述目标相似度范围指示的资源等级作为所述待推荐资源的资源等级；将所述待推荐资源的资源标识与所述资源等级对应存储，生成所述资源优先级列表。In another embodiment, the generating sub-module is configured to obtain a rating standard, where the rating standard includes a correspondence relationship between a plurality of resource levels and a similarity range; for each of the at least one resource to be recommended Resources to be recommended, to obtain the similarity of the resources to be recommended, to classify the similarities, to determine the target similarity range to which the similarity belongs; to determine the resource level indicated by the target similarity range as the to be recommended The resource level of the resource; storing the resource identifier of the resource to be recommended in correspondence with the resource level, and generating the resource priority list.

依据本申请第三方面，提供了一种计算机设备，包括存储器和处理器，所述存储器存储有计算机可读指令，所述处理器执行所述计算机可读指令时实现上述第一方面所述方法的步骤。According to a third aspect of the present application, there is provided a computer device including a memory and a processor, where the memory stores computer-readable instructions, and the processor implements the method described in the first aspect when executing the computer-readable instructions. A step of.

依据本申请第四方面，提供了一种计算机非易失性可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现第一方面所述的方法的步骤。According to a fourth aspect of the present application, a computer non-volatile readable storage medium is provided, on which computer-readable instructions are stored, and when the computer-readable instructions are executed by a processor, the method described in the first aspect is implemented. step.

借由上述技术方案，本申请提供的一种资源推荐方法、装置、计算机设备及计算机非易失性可读存储介质，与目前对用户画像进行分析的方式相比，本申请可以根据第一匹配规则和第二匹配规则确定待推荐用户，为待推荐用户确定至少一个待推荐资源，并基于至少一个待推荐资源对待推荐用户进行资源推荐，保证了对用户的喜好进行较为全面的分析，推荐给用户的资源准确，避免对推荐资源的浪费。By means of the above technical solution, a resource recommendation method, device, computer equipment and computer non-volatile readable storage medium provided by the present application, compared with the current method of analyzing user portraits, this application can be based on the first matching The rule and the second matching rule determine the user to be recommended, determine at least one resource to be recommended for the user to be recommended, and perform resource recommendation for the recommended user based on the at least one resource to be recommended, ensuring a comprehensive analysis of user preferences, and recommending to The user's resources are accurate to avoid wasting recommended resources.

上述说明仅是本申请技术方案的概述，为了能够更清楚了解本申请的技术手段，而可依照说明书的内容予以实施，并且为了让本申请的上述和其它目的、特征和优点能够更明显易懂，以下特举本申请的具体实施方式。The above description is only an overview of the technical solution of this application. In order to understand the technical means of this application more clearly, it can be implemented in accordance with the content of the description, and in order to make the above and other purposes, features, and advantages of this application more obvious and understandable. The specific implementations of this application are listed below.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

通过阅读下文优选实施方式的详细描述，各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。附图仅用于示出优选实施方式的目的，而并不认为是对本申请的限制。而且在整个附图中，用相同的参考符号表示相同的部件。在附图中：Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the detailed description of the preferred embodiments below. The drawings are only for the purpose of illustrating preferred embodiments and are not to be considered as limiting the present application. Moreover, the same reference numerals are used throughout the drawings to refer to the same parts. In the drawings:

图1示出了本申请实施例提供的一种资源推荐方法流程示意图；图2示出了本申请实施例提供的一种资源推荐方法流程示意图；图3A示出了本申请实施例提供的一种资源推荐装置的结构示意图；图3B示出了本申请实施例提供的一种资源推荐装置的结构示意图；图3C 示出了本申请实施例提供的一种资源推荐装置的结构示意图；图3D示出了本申请实施例提供的一种资源推荐装置的结构示意图。FIG. 1 shows a schematic flowchart of a resource recommendation method provided by an embodiment of the present application; FIG. 2 shows a schematic flowchart of a resource recommendation method provided by an embodiment of the present application; FIG. 3A shows a A structural schematic diagram of a resource recommendation device; FIG. 3B illustrates a schematic structural diagram of a resource recommendation device provided in an embodiment of the present application; FIG. 3C illustrates a schematic structural diagram of a resource recommendation device provided in an embodiment of the application; FIG. 3D A schematic structural diagram of a resource recommendation device according to an embodiment of the present application is shown.

具体实施方式detailed description

下面将参照附图更详细地描述本申请的示例性实施例。虽然附图中显示了本申请的示例性实施例，然而应当理解，可以以各种形式实现本申请而不应被这里阐述的实施例所限制。相反，提供这些实施例是为了能够更透彻地理解本申请，并且能够将本申请的范围完整的传达给本领域的技术人员。Hereinafter, exemplary embodiments of the present application will be described in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the application, and to fully convey the scope of the application to those skilled in the art.

本申请实施例提供了一种资源推荐方法，如图1所示，该方法包括：An embodiment of the present application provides a resource recommendation method. As shown in FIG. 1, the method includes:

101、根据第一匹配规则，确定多个候选用户，第一匹配规则至少包括目标年龄、目标地区和目标性别。其中，不同资源适用于的用户是不同的，只有针对用户的属性对用户进行资源推荐才可能实现用户对资源的查看及消费，如果盲目将与用户的属性不匹配的资源推荐给用户，反而会造成用户的方案，因此，获取至少一个用户的至少一个注册信息，将至少一个注册信息与第一匹配规则进行比对，并确定与第一匹配规则匹配的多个注册信息，将多个注册信息对应的多个用户作为多个候选用户，实现根据第一匹配规则，对用户的个人属性进行筛选，从而确定哪些资源应该推荐给哪些用户的过程，具体实施过程参见步骤201至步骤202中所示过程。101. Determine multiple candidate users according to a first matching rule. The first matching rule includes at least a target age, a target region, and a target gender. Among them, different resources are suitable for different users. Only by recommending resources to users based on their attributes can users view and consume resources. If they blindly recommend resources that do not match the user's attributes to users, they will instead A solution for causing a user, therefore, obtain at least one registration information of at least one user, compare at least one registration information with a first matching rule, and determine multiple registration information that matches the first matching rule, and compare multiple registration information Corresponding multiple users, as multiple candidate users, implement the process of filtering the user's personal attributes according to the first matching rule to determine which resources should be recommended to which users. The specific implementation process is shown insteps 201 to 202 process.

102、根据第二匹配规则，在多个候选用户中，提取至少一个待推荐用户，第二匹配规则为预设数目或预设忠诚度。发明人认识到，不同的用户进行资源推荐的价值是不同的，如果用户对资源感兴趣，并且很可能在后续对资源进行消费，则为用户进行资源推荐便是有价值的，否则是没有推荐价值的，因此，根据第二匹配规则，在多个候选用户中确定至少一个待推荐用户进行推荐，如果第二匹配规则为预设数目，则在多个候选用户中选取预设数目的候选用户作为至少一个待推荐用户；或如果第二匹配规则为预设忠诚度，则确定多个候选用户的多个用户忠诚度，将用户忠诚度与预设忠诚度匹配的候选用户作为至少一个待推荐用户。具体实施过程参见步骤203中所示过程。102. According to a second matching rule, at least one user to be recommended is extracted from a plurality of candidate users, and the second matching rule is a preset number or a preset loyalty. The inventor recognizes that the value of resource recommendation by different users is different. If the user is interested in the resource and is likely to consume the resource in the future, then the resource recommendation for the user is valuable, otherwise there is no recommendation It is valuable, therefore, according to the second matching rule, at least one user to be recommended is determined for recommendation from a plurality of candidate users. If the second matching rule is a preset number, a preset number of candidate users is selected from the plurality of candidate users. As at least one user to be recommended; or if the second matching rule is preset loyalty, determine multiple user loyalty of multiple candidate users, and use the candidate user whose user loyalty matches the preset loyalty as at least one to be recommended user. For a specific implementation process, refer to the process shown instep 203.

103、获取至少一个待推荐用户中每个待推荐用户的历史使用数据，历史使用数据至少包括搜索数据、浏览数据以及交易数据。在本申请实施例中，当确定至少一个待推荐用户后，由于每一个待推荐用户对产品方提供的资源的感兴趣程度是不同的，因此，对于至少一个待推荐用户中的每一个待推荐用户，均需要为该待推荐用户确定至少一个待推荐资源，具体实施过程参见步骤204中所示过程。103. Obtain historical usage data of each of the at least one user to be recommended. The historical usage data includes at least search data, browsing data, and transaction data. In the embodiment of the present application, after at least one user to be recommended is determined, since each user to be recommended has a different degree of interest in the resources provided by the product side, for each of the at least one user to be recommended, to be recommended The user needs to determine at least one resource to be recommended for the user to be recommended. For a specific implementation process, refer to the process shown instep 204.

104、确定应用的多个应用资源，采用预设算法，基于多个应用资源和历史使用数据，为至少一个待推荐用户中每个待推荐用户确定至少一个待推荐资源。在本申请实施例中，由于与用户的使用数据相似的产品资源用户可能会感兴趣，因此，当确定了用户的历史使用数据后，对应用提供的项目数据及产品数据进行统计，确定多个应用资源，对于至少一个待推荐用户中的每个待推荐用户，采用预设算法，计算多个应用资源与待推荐用户历史使用数据的多个相似度，并在多个应用资源中提取相似度大于预设相似度的至少一个应用资源作为至少一个待推荐资源，进而计算产品方提供的多个应用资源与历史使用数据的相似度，进而根据相似度为用户进行资源推荐，具体实施过程参见步骤205中所示过程。104. Determine multiple application resources of the application, and use a preset algorithm to determine at least one resource to be recommended for each of the at least one user to be recommended based on the multiple application resources and historical usage data. In the embodiment of the present application, since product resource users similar to the user's usage data may be of interest to users, after determining the user's historical usage data, the project data and product data provided by the application are counted to determine multiple Application resources. For each to-be-recommended user of at least one to-be-recommended user, a preset algorithm is used to calculate multiple similarities between multiple application resources and historical usage data of the to-be-recommended users, and extract similarities from multiple application resources. At least one application resource that is greater than the preset similarity is used as at least one resource to be recommended, and then the similarity between the multiple application resources provided by the product and the historical usage data is calculated, and then the user is recommended for resources based on the similarity. The process shown in 205.

105、对至少一个待推荐资源进行评级，并对推荐用户进行资源推荐。在本申请实施例中，当确定至少一个待推荐资源后，考虑到用户的终端可接受的推荐资源的数据量是有限的，大量的向用户进行资源推荐可能会对用户的正常使用造成影响，给用户带来困扰，因此，在确定至少一个待推荐资源后，可以基于预设模型算法，对至少一个待推荐资源进行评级，生成资源优先级列表，以便在后续对用户进行资源推荐时，先将优先级较高的资源推荐给用户，具体实施过程参见步骤206中所示过程。105. Rate at least one resource to be recommended, and recommend resources to the recommended user. In the embodiment of the present application, after determining at least one resource to be recommended, considering that the amount of data of the recommended resource that can be accepted by the user's terminal is limited, a large number of resource recommendations to the user may affect the normal use of the user. It causes confusion for users. Therefore, after determining at least one resource to be recommended, based on a preset model algorithm, the at least one resource to be recommended can be rated to generate a resource priority list, so that in the subsequent resource recommendation to the user, A resource with a higher priority is recommended to the user. For a specific implementation process, refer to the process shown instep 206.

本申请实施例提供的资源推荐方法，可以根据第一匹配规则和第二匹配规则确定待推荐用户，为待推荐用户确定至少一个待推荐资源，并基于至少一个待推荐资源对待推荐用户进行资源推荐，保证了对用户的喜好进行较为全面的分析，推荐给用户的资源准确，避免对推荐资源的浪费。The resource recommendation method provided in the embodiment of the present application may determine a user to be recommended according to the first matching rule and the second matching rule, determine at least one resource to be recommended for the user to be recommended, and perform resource recommendation to the recommended user based on the at least one resource to be recommended. , To ensure a more comprehensive analysis of user preferences, the resources recommended to users are accurate, to avoid wasting recommended resources.

本申请实施例提供了一种资源推荐方法，可以达到对用户的喜好进行较为全面的分析，推荐给用户的资源准确，避免对推荐资源的浪费的目的，如图2所示，该方法应用于手机、平板电脑等终端设备，该方法包括：The embodiment of the present application provides a resource recommendation method, which can achieve a more comprehensive analysis of user preferences, accurately recommend resources to users, and avoid waste of recommended resources. As shown in FIG. 2, the method is applied to Mobile phones, tablet computers and other terminal equipment, the method includes:

201、获取至少一个用户的至少一个注册信息，将至少一个注册信息与第一匹配规则进行比对。201. Acquire at least one registration information of at least one user, and compare the at least one registration information with a first matching rule.

发明人认识到，不同的用户在使用应用时，感兴趣的内容是不同的，如果应用的机制需要用户浏览大量资源后才能获取自身感兴趣的资源，则会导致用户的流失，使用该应用的用户会越来越少，因此，本申请基于用户本身的属性，为用户进行资源的推荐，保证用户获取感兴趣资源不至于复杂，满足不同用户的不同需求。其中，为了实现对用户进行的资源推荐，应用中可以搭载资源推荐系统，并基于该资源推荐系统实现对用户进行资源推荐。考虑到不同年龄、不同地区和不同性别的用户感兴趣的内容出入较为明显，因此，在对用户进行资源推荐时，可以获取用户的注册信息，以便后续确定是否向该用户进行资源推荐以及向该用户推荐哪些资源。应用中可以为用户提供注册入口，当检测到用户触发该注册入口时，显示注册页面，并当检测到用户对注册页面进行提交时，获取用户在注册页面输入的注册信息，将用户的注册信息存储，从而完成用户在应用中的注册，注册信息至少包括用户年龄、用户地区和用户性别。需要说明的是，由于应用中会存在大量注册的用户，为了对用户进行区分，便于对用户进行管理，在用户注册成功后，可以为用户分配注册账号，将该注册账号返回给用户，并将注册账号与用户的注册信息进行存储，使得后续用户可以基于注册账号登录到应用中，进而应用可以基于用户的注册信息为用户进行资源推荐。在实际应用中，由于不同的产品方提供给应用的资源的类型是不同的，某一个产品方仅能提供某一类型的资源，例如，对于食品的产品方，该产品方仅能提供有关食品方面的资源，因此，可以由产品方设置第一匹配规则，使得可以基于每个产品方提供的第一匹配规则，确定各个产品方可以进行资源推荐的候选用户，并在后续在这些候选用户中确定产品方的待推荐用户，使产品方可以向对其提供的资源感兴趣的用户进行资源推荐。其中，第一匹配规则至少包括目标年龄、目标地区和目标性别。需要说明的是，由于应用中提供的资源来自多个产品方，因此，每个产品方均可以提供自身的第一规则，以便应用的资源推荐系统根据该每个产品方提供的第一规则为每个产品方确定待推荐用户，进而为待推荐用户进行资源推荐。对于任一产品方来说，当获取到至少一个用户的至少一个注册信息后，将至少一个注册信息与该产品方提供的第一匹配规则进行比对，以便为该产品方确定多个候选用户。其中，在将至少一个注册信息与该产品方提供的第一匹配规则进行比对时，需要将注册信息中包括的全部内容与第一匹配规则包括的内容一一比对。The inventor recognizes that different users are different in the content of interest when using the application. If the mechanism of the application requires the user to browse a large number of resources in order to obtain the resources of interest, it will lead to the loss of users. There will be fewer and fewer users. Therefore, based on the user's own attributes, this application recommends resources for users to ensure that users' access to resources of interest is not complicated and meets the different needs of different users. Among them, in order to implement resource recommendation to users, an application may be equipped with a resource recommendation system, and based on the resource recommendation system, resource recommendation to users is implemented. Considering that users of different ages, regions, and genders are more interested in content, so when making resource recommendations to users, you can obtain user registration information to determine whether to make resource recommendations to the user and to the user. What resources users recommend. The application can provide a registration entry for the user. When the user detects that the registration entry is triggered, the registration page is displayed, and when the user submits the registration page, the registration information entered by the user on the registration page is obtained, and the user's registration information is obtained. Store to complete the registration of the user in the application. The registration information includes at least the user's age, user region, and user gender. It should be noted that, because there will be a large number of registered users in the application, in order to distinguish the users and facilitate the management of the users, after the user is successfully registered, the user can be assigned a registered account, the registered account is returned to the user, and The registered account and the user's registration information are stored, so that subsequent users can log in to the application based on the registered account, and the application can then recommend resources for the user based on the user's registration information. In practical applications, because the type of resources provided by different product parties is different, a certain product party can only provide a certain type of resource. For example, for a food product party, the product party can only provide relevant food products. Resources, therefore, the product side can set the first matching rule, so that based on the first matching rule provided by each product side, the candidate users that can be recommended for resources by each product side can be determined, and among these candidate users in the subsequent Identify the users to be recommended on the product side so that the product side can recommend resources to users who are interested in the resources they provide. The first matching rule includes at least a target age, a target region, and a target gender. It should be noted that, because the resources provided in the application come from multiple product parties, each product party can provide its own first rule, so that the resource recommendation system of the application according to the first rule provided by each product party is Each product side determines the users to be recommended, and then makes resource recommendations for the users to be recommended. For any product party, after obtaining at least one registration information of at least one user, the at least one registration information is compared with a first matching rule provided by the product party to determine multiple candidate users for the product party . When comparing at least one registration information with the first matching rule provided by the product side, it is necessary to compare all the content included in the registration information with the content included in the first matching rule.

202、确定与第一匹配规则匹配的多个注册信息，将多个注册信息对应的多个用户作为多个候选用户。202. Determine multiple pieces of registration information that match the first matching rule, and use multiple users corresponding to the multiple pieces of registration information as multiple candidate users.

在本申请实施例中，在将注册信息中包括的内容与第一匹配规则一一对比后，对于至少一个注册信息中的任一注册信息来说，当注册信息中包括的内容与第一匹配规则中的内容完全匹配时，则确定该注册信息与第一匹配规则匹配，这时，便可以将该注册信息对应的用户作为候选用户，进而基于至少一个注册信息，确定多个候选用户。例如，设当前存在注册信息A、注册信息B和注册信息C，注册信息中包括姓名A，年龄20岁，地区上海，性别男；注册信息B中包括姓名B，年龄50岁，地区上海，性别男；注册信息C中包括姓名C、年龄21岁，地区北京，性别男，如果第一匹配规则中包括的内容为目标年龄18岁至20岁，目标地区为上海，目标性别为男，则上述注册信息A、B和C中，只有注册信息A完成与第一匹配规则匹配，因此，可以将注册信息A对应的用户作为候选用户。另外，只要注册信息中的任一项内容与第一匹配规则不匹配，则该注册信息对应的用户便不能作为候选用户。需要说明的是，对于至少一个用户的至少一个注册信息，均可以通过上述步骤确定注册信息是否与第一匹配规则匹配。In the embodiment of the present application, after comparing the content included in the registration information with the first matching rule one by one, for any registration information in the at least one registration information, when the content included in the registration information matches the first When the content in the rule completely matches, it is determined that the registration information matches the first matching rule. At this time, the user corresponding to the registration information can be used as a candidate user, and then multiple candidate users are determined based on at least one registration information. For example, suppose that there is currently registration information A, registration information B, and registration information C. The registration information includes name A, age 20 years old, region Shanghai, gender male; registration information B includes name B, age 50 years old, region Shanghai, gender Male; Registration information C includes name C, age 21, region Beijing, gender male, if the content included in the first matching rule is target age 18 to 20, target region is Shanghai, and target gender is male, then the above Among the registration information A, B, and C, only the registration information A completes the matching with the first matching rule. Therefore, the user corresponding to the registration information A can be used as a candidate user. In addition, as long as any content in the registration information does not match the first matching rule, the user corresponding to the registration information cannot be a candidate user. It should be noted that, for at least one registration information of at least one user, whether the registration information matches the first matching rule can be determined through the foregoing steps.

203、根据第二匹配规则，在多个候选用户中，提取至少一个待推荐用户，第二匹配规则至少为预设数目或预设忠诚度。203. According to a second matching rule, at least one user to be recommended is extracted from a plurality of candidate users, and the second matching rule is at least a preset number or a preset loyalty.

在本申请实施例中，由于应用中进行注册的用户的用户数量较大，使得根据第一匹配规则进行筛选后确定的候选用户的用户数量相对也会较大，考虑到应用进行资源推荐的能力是有限的，为过量的用户进行资源推荐会给应用造成压力，影响应用的正常运行，因此，可以设置第二匹配规则，并在确定多个候选用户后，根据第二匹配规则，在多个候选用户中提取至少一个待推荐用户，以便在后续仅为至少一个待推荐用户进行资源推荐，从而减轻应用的资源推荐压力。其中，第二匹配规则具体可为预设数目或预设忠诚度中的一种，可由产品方定制，如果产品方希望基于数量对用户进行资源推荐，则可定制预设数目作为第二匹配规则，具体地，可以根据应用推荐资源的能力设置预设数目；如果产品方希望将资源推荐给经常关注自己产品的用户，也即希望基于用户对产品的忠诚度对用户进行资源推荐，则可定制预设忠诚度作为第二匹配规则。本申请实施例对于第二匹配规则的具体内容不进行限定。相应地，如果第二匹配规则为预设数目，则在多个候选用户中选取预设数目的候选用户作为至少一个待推荐用户，在选取至少一个待推荐用户时，可以随机进行选取。例如，设第二匹配规则中的预设数目为1万，如果确定的候选用户为10万人，则在10万个候选用户中，随机选取1万人作为待推荐用户。具体地，在进行随机选取时，可以按照确定候选用户的时间顺序为每一个候选用户进行编号，给每一个候选用户生成一个用户编号，按顺序选取预设数目的用户编号中包括特定数字的用户作为待推荐用户，例如，设将用户编号中包括8的候选用户作为待推荐用户；或者按顺序选取编号为偶数的候选用户作为待推荐用户等，本申请实施例对选取待推荐用户的方式不进行具体限定。如果第二匹配规则为预设忠诚度，则确定多个候选用户的多个用户忠诚度，将用户忠诚度与预设忠诚度匹配的候选用户作为至少一个待推荐用户。其中，在确定每一个候选用户的忠诚度时，可以统计每一个候选用户在应用中的支出金额，将支出金额作为每个候选用户产品的忠诚度；或者统计每一个候选用户使用应用的总时长，将总时长作为每个候选用户对该应用的忠诚度。当确定每个候选用户的忠诚度后，将忠诚度大于预设忠诚度的候选用户作为待推荐用户。本申请实施例对确定候选用户的忠诚度的方式不进行具体限定。In the embodiment of the present application, since the number of registered users in the application is large, the number of candidate users determined after filtering according to the first matching rule is also relatively large. Considering the ability of the application to perform resource recommendation It is limited, and recommending resources for excessive users will put pressure on the application and affect the normal operation of the application. Therefore, a second matching rule can be set, and after determining multiple candidate users, according to the second matching rule, At least one user to be recommended is extracted from the candidate users, so that resource recommendation is performed only for the at least one user to be recommended in the future, thereby reducing the resource recommendation pressure of the application. The second matching rule may be one of a preset number or preset loyalty, which may be customized by the product side. If the product side wishes to recommend resources to users based on the number, the preset number may be customized as the second matching rule. Specifically, the preset number can be set according to the ability of the application to recommend resources; if the product side wants to recommend resources to users who often follow their products, that is, they want to recommend resources to users based on their loyalty to the product, it can be customized Preset loyalty as the second matching rule. The embodiment of the present application does not limit the specific content of the second matching rule. Correspondingly, if the second matching rule is a preset number, a preset number of candidate users are selected as the at least one user to be recommended from among a plurality of candidate users, and when the at least one user to be recommended is selected, the selection may be performed randomly. For example, suppose the preset number in the second matching rule is 10,000. If the determined candidate user is 100,000, among the 100,000 candidate users, 10,000 are randomly selected as users to be recommended. Specifically, when random selection is performed, each candidate user may be numbered in a chronological order in which the candidate users are determined, a user number may be generated for each candidate user, and a preset number of user numbers including a specific number of users may be sequentially selected. As a user to be recommended, for example, a candidate user whose user number includes 8 is set as the user to be recommended; or a candidate user with an even number is sequentially selected as the user to be recommended. The embodiment of the present application does not change the method of selecting the user to be recommended. Be specifically limited. If the second matching rule is preset loyalty, multiple user loyalty of multiple candidate users is determined, and the candidate user whose user loyalty matches the preset loyalty is at least one user to be recommended. Among them, when determining the loyalty of each candidate user, the amount of expenditure of each candidate user in the application can be counted, and the amount of expenditure can be used as the loyalty of each candidate user's product; or the total length of time that each candidate user uses the application is counted , Use the total time as the loyalty of each candidate user to the app. After determining the loyalty of each candidate user, the candidate user whose loyalty is greater than the preset loyalty is regarded as the user to be recommended. The embodiment of the present application does not specifically limit the manner of determining the loyalty of the candidate user.

204、获取至少一个待推荐用户中每个待推荐用户的历史使用数据。204. Obtain historical usage data of each to-be-recommended user of the at least one to-be-recommended user.

考虑到用户在使用应用时，会对感兴趣的资源重点关注，进而在应用中搜索感兴趣的内容，浏览感兴趣的内容，并对感兴趣的产品进行购买，因此，可以将用户搜索过的关键字、用户浏览过的网页中包括的内容以及购买过的产品的购买产品信息作为用户的历史使用数据，也即将搜索数据、浏览数据以及交易数据作为历史使用数据，以便在后续根据历史使用数据为用户推荐其兴趣度较高的资源。需要说明的是，考虑到用户的好友的感兴趣的内容可能会对用户的感兴趣的资源造成影响，使得用户可能也对其好友感兴趣的内容感兴趣，因此，历史使用数据中还可以包括用户的好友购买过产品的产品信息，本申请实施例对历史使用数据包括的内容不进行具体限定。Considering that when users use the application, they will pay attention to the resources they are interested in, and then search for the content of interest in the application, browse the content of interest, and make purchases of the products that interest them. Keywords, content included in webpages viewed by users, and purchased product information of purchased products are used as historical user data, and search data, browsing data, and transaction data are also used as historical usage data, in order to use the data in the future based on history Recommend resources with high interest for users. It should be noted that considering that the user's friend's interesting content may affect the user's interesting resources, so that the user may also be interested in the content that his friend is interested in. Therefore, the historical usage data may also include The product information of the product that the friend of the user has purchased, the embodiment of the present application does not specifically limit the content included in the historical usage data.

205、对于所述至少一个待推荐用户中的每个待推荐用户，采用所述预设算法，计算所述多个应用资源与所述待推荐用户历史使用数据的多个相似度，在所述多个应用资源中提取相似度大于预设相似度的至少一个应用资源作为所述至少一个待推荐资源。205. For each of the at least one user to be recommended, the preset algorithm is used to calculate a plurality of similarities between the multiple application resources and the historical use data of the user to be recommended. At least one application resource with a similarity greater than a preset similarity is extracted from the multiple application resources as the at least one resource to be recommended.

在计算多个应用资源与历史使用数据的多个相似度时，可为分别为关键字、网页内容和购买产品信息设置权重，对于多个应用资源中的每个应用资源，采用预设算法，计算该应用资源所包括的关键字的个数在应用资源包括的全部内容中所占的第一比例，计算应用资源与网页内容之间相同的内容在应用资源包括的全部内容中所占的第二比例，计算应用资源的产品的功能与购买信息中涉及的产品具有的相同功能在应用资源的产品具有的总功能中所占的第三比例；最后，按照关键字、网页内容和购买产品信息的权重，将第一比例、第二比例和第三比例结合起来，得到该应用资源与历史使用数据的相似度。例如，设关键字的权重为40％，网页内容的权重为10％，购买产品信息的权重为50％，如果计算得到应用资源A的第一比例、第二比例和第三比例分别为98％、34％以及60％，则可以确定应用资源A与历史使用数据之间的相似度为98％×40％+34％×10％+60％×50％，也即应用资源A与历史使用数据之间的相似度为72.6％。When calculating multiple similarities between multiple application resources and historical usage data, you can set weights for keywords, web content, and purchased product information separately. For each application resource in the multiple application resources, a preset algorithm is used. Calculate the first proportion of the number of keywords included in the application resource among all the content included in the application resource, and calculate the first proportion of the same content between the application resource and the web page content in all the content included in the application resource. The second ratio is to calculate the third ratio of the functions of the products of the application resources and the products involved in the purchase information to the total functions of the products of the application resources; finally, according to the keywords, web content and purchase product information Combining the first ratio, the second ratio, and the third ratio to obtain the similarity between the application resource and the historical usage data. For example, set the weight of keywords to 40%, the weight of web content to 10%, and the weight of purchased product information to 50%. If calculated, the first, second, and third proportions of application resource A are 98%. , 34%, and 60%, it can be determined that the similarity between the application resource A and the historical usage data is 98% × 40% + 34% × 10% + 60% × 50%, that is, the application resource A and the historical usage data The similarity between them is 72.6%.

通过计算多个应用资源中的每一个应用资源与历史使用数据之间的相似度，可以得到多个相似度，由于相似度越高表示应用资源与历史使用数据越相似，用户对过小相似度对应的应用资源感兴趣的可能性很低，因此，可以设置一个用于对相似度进行筛选的标准作为预设相似度，将大于预设相似度对应的应用资源提取出来作为至少一个待推荐资源，进而保证至少一个待推荐资源满足用户的兴趣要求。例如，设预设相似度为80％，如果确定应用资源A与历史使用数据的相似度为60％，应用资源B与历史使用数据的相似度为90％，应用资源C与历史使用数据的相似度为95％，则可将应用资源B和应用资源C作为待推荐资源。By calculating the similarity between each of the multiple application resources and the historical usage data, multiple similarities can be obtained. Since the higher the similarity, the more similar the application resources and historical usage data are, the less similarity the user has. The possibility of corresponding application resources is very low, so you can set a criterion for filtering similarity as the preset similarity, and extract the application resources corresponding to the preset similarity as at least one resource to be recommended. , Thereby ensuring that at least one resource to be recommended meets the user's interest requirements. For example, if the preset similarity is 80%, if it is determined that the similarity between application resource A and historical usage data is 60%, the similarity between application resource B and historical usage data is 90%, and the similarity between application resource C and historical usage data If the degree is 95%, the application resource B and the application resource C can be used as the resources to be recommended.

需要说明的是，预设算法可为权重算法、逻辑回归算法、神经网络算法，本申请实施例以预设算法为权重算法进行说明，本申请实施例对预设算法具体为哪一种算法不进行具体限定。It should be noted that the preset algorithm may be a weight algorithm, a logistic regression algorithm, or a neural network algorithm. In the embodiment of the present application, the preset algorithm is used as the weight algorithm. The embodiment of the present application does not specifically describe which algorithm the preset algorithm is. Be specifically limited.

206、对所述至少一个待推荐资源进行分类，为所述至少一个待推荐资源评级，生成资源优先级列表，将所述资源优先级列表中资源等级大于预设等级的待推荐资源推荐给所述推荐用户。206. Classify the at least one resource to be recommended, rank the at least one resource to be recommended, generate a resource priority list, and recommend the resource to be recommended in the resource priority list with a resource level greater than a preset level to the resource. Said recommended users.

其中，在对至少一个待推荐资源进行评级时，可以通过执行下述步骤一至步骤三中的过程实现。Wherein, when the at least one resource to be recommended is rated, it can be implemented by performing the following steps from step 1 to step 3.

步骤一、获取评级标准，评级标准包括多个资源等级与相似度范围之间的对应关系。Step 1: Obtain a rating standard. The rating standard includes a correspondence relationship between multiple resource levels and similarity ranges.

为了对待推荐资源进行评级，可以设置评级标准，规定各个资源等级对应的相似度范围。具体地，可以设置第一阈值、第二阈值和第三阈值，将相似度大于第一阈值的相似度对应的待推荐资源作为A级，将相似度大于第二阈值小于第一阈值的相似度对应的待推荐资源作为 B级，将相似度大于第三阈值小于第二阈值的相似度对应的待推荐资源作为C级，将相似度小于第三阈值的相似度对应的待推荐资源作为D级。例如，将相似度大于90％的待推荐资源作为A级，将相似度大于70％小于90％的待推荐资源作为B级，将相似度大于50％小于70％的待推荐资源作为C级，将相似度小于50％的待推荐资源作为D级。In order to rate the recommended resources, a rating standard can be set to specify the similarity range corresponding to each resource level. Specifically, a first threshold, a second threshold, and a third threshold may be set, and a resource to be recommended corresponding to a similarity greater than the first threshold is regarded as a level A, and the similarity greater than the second threshold is smaller than the first threshold. The corresponding resource to be recommended is regarded as level B, the resource to be recommended corresponding to the degree of similarity greater than the third threshold and smaller than the second threshold is regarded as C, and the resource to be recommended corresponding to the similarity less than the third threshold is regarded as D. . For example, a resource to be recommended with similarity greater than 90% is classified as A, a resource to be recommended with similarity greater than 70% and less than 90% is classified as B, and a resource to be recommended with similarity greater than 50% and less than 70% is classified as C, A resource to be recommended with a similarity of less than 50% is regarded as a D-level.

步骤二、对于所述至少一个待推荐资源中的每个待推荐资源，获取所述待推荐资源的相似度，对所述相似度进行分类，确定所述相似度所属的目标相似度范围。Step 2: For each to-be-recommended resource of the at least one to-be-recommended resource, obtain the similarity of the to-be-recommended resource, classify the similarity, and determine a target similarity range to which the similarity belongs.

当获取了评级标准后，对至少一个待推荐资源的相似度进行划分，确定相似度所属的目标相似度范围。继续以上述步骤一中的例子为例，如果应用资源1的相似度为98％，则应用资源1的相似度在相似度大于90％的范围内；如果应用资源2的相似度为88％，则应用资源2的相似度在大于70％小于90％的范围内；如果应用资源3的相似度为66％，则应用资源3的相似度在大于50％小于70％的范围内。After obtaining the rating criteria, the similarity of at least one resource to be recommended is divided to determine the target similarity range to which the similarity belongs. Continue to take the example in step 1 as an example. If the similarity of application resource 1 is 98%, the similarity of application resource 1 is within the range of greater than 90%; if the similarity of application resource 2 is 88%, Then the similarity of the application resource 2 is within a range of greater than 70% and less than 90%; if the similarity of the application resource 3 is 66%, the similarity of the application resource 3 is within a range of greater than 50% and less than 70%.

步骤三、确定所述目标相似度范围指示的资源等级作为所述待推荐资源的资源等级。Step 3: Determine the resource level indicated by the target similarity range as the resource level of the resource to be recommended.

当确定了目标相似度范围后，便可以将目标相似度范围指示的资源等级作为待推荐资源的资源等级，实现为待推荐资源确定资源等级。继续以上述步骤一中的例子为例，如果应用资源1的相似度为98％，则应用资源1的资源等级为A级；如果应用资源2的相似度为88％，则应用资源2的资源等级为B级；如果应用资源3的相似度为66％，则应用资源3的资源等级为C级；如果应用资源4的相似度为44％，则应用资源4的资源等级为D级。After the target similarity range is determined, the resource level indicated by the target similarity range can be used as the resource level of the resource to be recommended, and the resource level can be determined for the resource to be recommended. Continue to use the example in step 1 as an example. If the similarity of application resource 1 is 98%, the resource level of application resource 1 is A; if the similarity of application resource 2 is 88%, the resource of application resource 2 is applied. The level is B; if the similarity of application resource 3 is 66%, the resource level of application resource 3 is C; if the similarity of application resource 4 is 44%, the resource level of application resource 4 is D.

步骤四、将待推荐资源的资源标识与资源等级对应存储，生成资源优先级列表。Step 4. Store the resource identifier of the resource to be recommended in correspondence with the resource level, and generate a resource priority list.

当确定至少一个待推荐资源的资源等级后，便可以将待推荐资源的资源标识与资源等级对应存储，进而生成资源优先级列表。需要说明的是，在将待推荐资源的资源标识与资源等级对应存储时，还可以对应存储待推荐资源的相似度，以便在进行资源推荐时，还可以将相似度展示给待推荐用户。After determining the resource level of at least one resource to be recommended, the resource identifier of the resource to be recommended and the resource level may be stored in correspondence, and a resource priority list may be generated. It should be noted that, when the resource identifier of the resource to be recommended is correspondingly stored with the resource level, the similarity of the resource to be recommended may also be stored correspondingly, so that when the resource recommendation is performed, the similarity may also be displayed to the user to be recommended.

当生成至少一个待推荐资源的资源优先级列表后，便可以按照资源优先级列表为待推荐用户推荐资源。其中，可以设置将哪些资源等级的应用资源推荐给待推荐用户，并将该资源等级对应的全部应用资源推荐给待推荐用户。例如，如果应用中设置将资源等级为A级和B级的待推荐资源推荐给待推荐用户，则将资源等级为A级和B级中的待推荐资源推荐给待推荐用户。After generating a resource priority list of at least one resource to be recommended, a resource can be recommended for the user to be recommended according to the resource priority list. Wherein, it is possible to set which resource level application resources are recommended to the user to be recommended, and to recommend all application resources corresponding to the resource level to the user to be recommended. For example, if it is set in the application to recommend resources to be recommended to users to be recommended with resource levels A and B, the resources to be recommended in resource levels A and B are recommended to users to be recommended.

在实际应用的过程中，考虑到随着时间的变化，待推荐用户的兴趣爱好也会发生变化，因此，资源推荐系统中可以设置预设周期，每隔预设周期，执行确定至少一个待推荐用户并为至少一个待推荐用户中每个待推荐用户进行资源推荐的过程，在保证产品方进行资源推荐的待推荐用户一定是对产品方提供的产品感兴趣的用户的同时，还保证推荐给待推荐用户的资源一定是待推荐用户感兴趣的，避免用户的喜好发生变化时，仍然给用户推荐以前喜欢的内容，造成用户的反感。In the actual application process, in consideration of changes in time, the interests of users to be recommended will also change. Therefore, a preset period can be set in the resource recommendation system, and every predetermined period is executed to determine at least one to be recommended. The process of users making resource recommendations for each of the at least one to-be-recommended users. While ensuring that the to-be-recommended users on the product side must be users who are interested in the products provided by the product side, it also guarantees the recommendation to The resources of the user to be recommended must be of interest to the user to be recommended. When the user's preferences are changed, the user still recommends the previously liked content, which causes the user's dislike.

进一步地，作为图1所述方法的具体实现，本申请实施例提供了一种资源推荐装置，如图3A所示，所述装置包括：第一确定模块301，提取模块302，获取模块303，第二确定模块304和推荐模块305。Further, as a specific implementation of the method described in FIG. 1, an embodiment of the present application provides a resource recommendation device. As shown in FIG. 3A, the device includes afirst determination module 301, anextraction module 302, and anacquisition module 303. Thesecond determination module 304 and therecommendation module 305.

该第一确定模块301，用于根据第一匹配规则，确定多个候选用户，第一匹配规则至少包括目标年龄、目标地区和目标性别；The first determiningmodule 301 is configured to determine a plurality of candidate users according to a first matching rule, where the first matching rule includes at least a target age, a target region, and a target gender;

该提取模块302，用于根据第二匹配规则，在多个候选用户中，提取至少一个待推荐用户，第二匹配规则为预设数目或预设忠诚度；Theextraction module 302 is configured to extract at least one user to be recommended from a plurality of candidate users according to a second matching rule, where the second matching rule is a preset number or a preset loyalty;

该获取模块303，用于获取至少一个待推荐用户中每个待推荐用户的历史使用数据，历史使用数据至少包括搜索数据、浏览数据以及交易数据；The obtainingmodule 303 is configured to obtain historical usage data of each to-be-recommended user among at least one to-be-recommended user. The historical use data includes at least search data, browsing data, and transaction data.

该第二确定模块304，用于获取至少一个待推荐用户中每个待推荐用户的历史使用数据，历史使用数据至少包括搜索数据、浏览数据以及交易数据；The second determiningmodule 304 is configured to obtain historical usage data of each to-be-recommended user among at least one to-be-recommended user. The historical use data includes at least search data, browsing data, and transaction data.

该推荐模块305，用于确定应用的多个应用资源，采用预设算法，基于多个应用资源和历史使用数据，为至少一个待推荐用户中每个待推荐用户确定至少一个待推荐资源。Therecommendation module 305 is configured to determine multiple application resources of an application, and uses a preset algorithm to determine at least one resource to be recommended for each of the at least one user to be recommended based on the multiple application resources and historical usage data.

在具体的应用场景中，如图3B所示，该第一确定模块301，包括比对子模块3011和确定子模块3012。In a specific application scenario, as shown in FIG. 3B, thefirst determination module 301 includes acomparison sub-module 3011 and adetermination sub-module 3012.

该比对子模块3011，用于获取至少一个用户的至少一个注册信息，将至少一个注册信息与第一匹配规则进行比对，注册信息至少包括用户年龄、用户地区和用户性别；The comparison sub-module 3011 is configured to obtain at least one registration information of at least one user, and compare the at least one registration information with a first matching rule. The registration information includes at least a user age, a user region, and a user gender.

该确定子模块3012，用于确定与第一匹配规则匹配的多个注册信息，将多个注册信息对应的多个用户作为多个候选用户。The determining sub-module 3012 is configured to determine a plurality of registration information matching the first matching rule, and regard a plurality of users corresponding to the plurality of registration information as a plurality of candidate users.

在具体的应用场景中，该提取模块302，用于如果第二匹配规则为预设数目，则在多个候选用户中选取预设数目的候选用户作为至少一个待推荐用户；或，如果第二匹配规则为预设忠诚度，则确定多个候选用户的多个用户忠诚度，将用户忠诚度与预设忠诚度匹配的候选用户作为至少一个待推荐用户。In a specific application scenario, theextraction module 302 is configured to select a preset number of candidate users as at least one user to be recommended if the second matching rule is a preset number; or, if the second matching rule is a preset number; If the matching rule is preset loyalty, multiple user loyalty of multiple candidate users is determined, and the candidate user whose user loyalty matches the preset loyalty is regarded as at least one user to be recommended.

在具体的应用场景中，如图3C所示，该第二确定模块304，包括获取子模块3041，计算子模块3042和确定子模块3043。In a specific application scenario, as shown in FIG. 3C, thesecond determination module 304 includes anacquisition submodule 3041, acalculation submodule 3042, and adetermination submodule 3043.

该获取子模块3041，用于对应用提供的项目数据及产品数据进行统计，确定多个应用资源；The acquisition submodule 3041 is configured to perform statistics on project data and product data provided by the application, and determine multiple application resources;

该计算子模块3042，用于对于至少一个待推荐用户中的每个待推荐用户，采用预设算法，计算多个应用资源与待推荐用户历史使用数据的多个相似度；Thecalculation submodule 3042 is configured to calculate a plurality of similarities between multiple application resources and historical usage data of the users to be recommended by using a preset algorithm for each of the at least one user to be recommended;

该确定子模块3043，用于在多个应用资源中提取相似度大于预设相似度的至少一个应用资源作为至少一个待推荐资源。The determining sub-module 3043 is configured to extract at least one application resource with a similarity greater than a preset similarity from a plurality of application resources as at least one resource to be recommended.

在具体的应用场景中，该计算子模块3042，用于对于多个应用资源中的任一应用资源，采用预设算法，计算搜索数据在应用资源的内容中所占的第一比例，计算浏览数据在应用资源的内容中所占的第二比例，计算交易数据在应用资源的内容中所占的第三比例；分别确定搜索数据的第一权重、浏览数据的第二权重以及交易数据的第三权重；计算第一比例与第一权重的第一乘积，计算第二比例与第二权重的第二乘积，计算第三比例与第三权重的第三乘积；获取第一乘积、第二乘积以及第三乘积的和值，将和值作为应用资源与历史使用数据的相似度。In a specific application scenario, thecalculation submodule 3042 is configured to use a preset algorithm for any application resource among multiple application resources, calculate a first proportion of search data in the content of the application resource, and calculate a browse The second proportion of data in the content of the application resource, calculates the third proportion of transaction data in the content of the application resource; determining the first weight of the search data, the second weight of the browsing data, and the Three weights; calculate the first product of the first proportion and the first weight, calculate the second product of the second proportion and the second weight, calculate the third product of the third proportion and the third weight; obtain the first product and the second product And the sum of the third product, and the sum is used as the similarity between the application resource and the historical usage data.

在具体的应用场景中，如图3D所示，该推荐模块305，包括生成子模块3051和推荐子模块3052。In a specific application scenario, as shown in FIG. 3D, therecommendation module 305 includes ageneration sub-module 3051 and arecommendation sub-module 3052.

该生成子模块3051，用于对至少一个待推荐资源进行分类，为至少一个待推荐资源评级，生成资源优先级列表；The generating sub-module 3051 is configured to classify at least one resource to be recommended, generate a resource priority list for the at least one resource to be recommended;

该推荐子模块3052，用于将资源优先级列表中资源等级大于预设等级的待推荐资源推荐给推荐用户。Therecommendation submodule 3052 is configured to recommend a resource to be recommended in a resource priority list with a resource level greater than a preset level to a recommending user.

在具体的应用场景中，该生成子模块3051，用于获取评级标准，评级标准包括多个资源等级与相似度范围之间的对应关系；对于至少一个待推荐资源中的每个待推荐资源，获取待推荐资源的相似度，对相似度进行分类，确定相似度所属的目标相似度范围；确定目标相似度范围指示的资源等级作为待推荐资源的资源等级；将待推荐资源的资源标识与资源等级对应存储，生成资源优先级列表。In a specific application scenario, the generating sub-module 3051 is configured to obtain a rating standard, where the rating standard includes a correspondence relationship between multiple resource levels and similarity ranges; for each resource to be recommended in at least one resource to be recommended, Obtain the similarity of the resource to be recommended, classify the similarity, and determine the target similarity range to which the similarity belongs; determine the resource level indicated by the target similarity range as the resource level of the resource to be recommended; use the resource identifier and resource of the resource to be recommended The level corresponds to the storage, and a resource priority list is generated.

需要说明的是，本申请实施例提供的一种资源推荐装置所涉及各功能单元的其他相应描述，可以参考图1中的对应描述，在此不再赘述。It should be noted that, for other corresponding descriptions of the functional units involved in the resource recommendation device provided in the embodiment of the present application, reference may be made to the corresponding description in FIG. 1, and details are not described herein again.

基于上述如图1所示方法，相应的，本申请实施例还提供了一种存储设备，其上存储有计算机可读指令，该程序被处理器执行时实现上述如图1所示的资源推荐方法。Based on the method shown in FIG. 1, correspondingly, an embodiment of the present application further provides a storage device that stores computer-readable instructions. When the program is executed by a processor, the resource recommendation shown in FIG. 1 is implemented. method.

基于上述如图1所示方法和如图3A至图3D所示虚拟装置的实施例，为了实现上述目的，本申请实施例还提供了一种资源推荐的实体装置，该实体装置包括存储设备和处理器；所述存储设备，用于存储计算机可读指令；所述处理器，用于执行所述计算机可读指令以实现上述如图1所示的资源推荐方法。Based on the method shown in FIG. 1 and the embodiments of the virtual device shown in FIG. 3A to FIG. 3D, in order to achieve the above purpose, an embodiment of the present application further provides a resource recommendation physical device, which includes a storage device and A processor; the storage device for storing computer-readable instructions; the processor for executing the computer-readable instructions to implement the resource recommendation method shown in FIG. 1.

通过应用本申请的技术方案，可以根据第一匹配规则和第二匹配规则确定待推荐用户，为待推荐用户确定至少一个待推荐资源，并根据预设模型算法，基于至少一个待推荐资源对待推荐用户进行资源推荐，保证了对用户的喜好进行较为全面的分析，推荐给用户的资源准确，避免对推荐资源的浪费。By applying the technical solution of this application, it is possible to determine a user to be recommended according to the first matching rule and the second matching rule, determine at least one resource to be recommended for the user to be recommended, and treat the recommendation based on the at least one resource to be recommended according to a preset model algorithm The user's resource recommendation ensures a more comprehensive analysis of the user's preferences, and the resources recommended to the user are accurate, avoiding waste of recommended resources.

通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到本申请可以通过硬件实现，也可以借助软件加必要的通用硬件平台的方式来实现。基于这样的理解，本申请的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施场景所述的方法。本领域技术人员可以理解附图只是一个优选实施场景的示意图，附图中的模块或流程并不一定是实施本申请所必须的。本领域技术人员可以理解实施场景中的装置中的模块可以按照实施场景描述进行分布于实施场景的装置中，也可以进行相应变化位于不同于本实施场景的一个或多个装置中。上述实施场景的模块可以合并为一个模块，也可以进一步拆分成多个子模块。上述本申请序号仅仅为了描述，不代表实施场景的优劣。以上公开的仅为本申请的几个具体实施场景，但是，本申请并非局限于此，任何本领域的技术人员能思之的变化都应落入本申请的保护范围。Through the description of the above embodiments, those skilled in the art can clearly understand that this application can be implemented by hardware, or by software plus necessary general hardware platform. Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.), including several The instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of this application. Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required to implement this application. Those skilled in the art can understand that the modules in the device in the implementation scenario may be distributed among the devices in the implementation scenario according to the description of the implementation scenario, or may be correspondingly located in one or more devices different from the implementation scenario. The modules of the above implementation scenario can be combined into one module, or further divided into multiple sub-modules. The above serial number of this application is only for description, and does not represent the advantages and disadvantages of the implementation scenario. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to this, and any changes that can be thought by those skilled in the art should fall into the protection scope of this application.