技术领域Technical Field
本发明涉及计算机技术领域，特别涉及人工智能技术领域，尤其涉及一种产品推荐方法和装置。The present invention relates to the field of computer technology, in particular to the field of artificial intelligence technology, and more particularly to a product recommendation method and device.
背景技术Background Art
目前各大银行的网站或应用程序中理财产品申购页面中都会有理财产品推荐的模块，而在这些模块中所选的产品主要是由银行选出，或是根据过往的购买记录结合用户特征所预测当前用户可能会购买的产品。相关技术中，由银行选出的产品通常是根据其业绩、购买人数等指标进行筛选，这种方式推荐的产品数量有限且推荐精度较低，用户体验较差；预测当前用户的购买情况则需要对较大范围购买过理财产品的用户进行训练，对于新用户而言，银行无法准确的获取他们的全部信息从而使得银行无法精准的推荐产品，往往只能给新用户推荐银行自选产品。而且推荐给老用户的产品也是基于其过往投资经历，而无法灵活适应其将来的投资计划，推荐准确度较低，从而导致销售效率低下。At present, there are modules for recommending financial products in the financial product subscription pages on the websites or applications of major banks. The products selected in these modules are mainly selected by the bank, or are predicted to be the products that the current user may purchase based on past purchase records and user characteristics. In related technologies, the products selected by the bank are usually screened based on indicators such as their performance and the number of purchasers. This method recommends a limited number of products and has low recommendation accuracy, resulting in poor user experience. Predicting the current user's purchase situation requires training a large range of users who have purchased financial products. For new users, the bank cannot accurately obtain all their information, making it impossible for the bank to accurately recommend products. It can only recommend bank-selected products to new users. Moreover, the products recommended to old users are also based on their past investment experience, and cannot flexibly adapt to their future investment plans. The recommendation accuracy is low, resulting in low sales efficiency.
发明内容Summary of the invention
本发明的一个目的在于提供一种产品推荐方法，对已购产品的交易数据对用户进行聚类分析，结合问卷应答以无侵入的方式嵌入原有的产品申购页面，通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，结合用户偏好和产品收益，更加精准的预测用户的购买产品，更加快捷的促成交易，提高产品推荐精度，从而提升用户体验，进而提升产品的销售效率。本发明的另一个目的在于提供一种产品推荐装置。本发明的再一个目的在于提供一种计算机可读介质。本发明的还一个目的在于提供一种计算机设备。One object of the present invention is to provide a product recommendation method, which performs cluster analysis on the transaction data of purchased products, embeds the data into the original product purchase page in a non-invasive manner in combination with questionnaire answers, accurately classifies users by improving user characteristics and updating user information in real time, and more accurately predicts the user's purchase products in combination with user preferences and product benefits, facilitates transactions more quickly, improves product recommendation accuracy, thereby improving user experience, and further improving product sales efficiency. Another object of the present invention is to provide a product recommendation device. Another object of the present invention is to provide a computer-readable medium. Still another object of the present invention is to provide a computer device.
为了达到以上目的，本发明一方面公开了一种产品推荐方法，包括：In order to achieve the above objectives, the present invention discloses a product recommendation method, comprising:
获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；Obtain transaction data of purchased products, including product information and user information of transaction users of purchased products;
通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；Through clustering algorithm, transaction users are clustered according to user information to obtain user clustering results;
根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；Generate pre-recommended product information corresponding to the user type based on the user clustering results, product information, and user information of the transaction users who have purchased the product;
通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息。Through the classification algorithm, according to the user clustering results and the pre-recommended product information corresponding to the user type, the pre-generated user portrait of the target user is classified and predicted to generate the target recommended product information.
优选的，获取已购产品的交易数据，包括：Preferably, obtaining transaction data of purchased products includes:
获取指定时间段内已购产品的产品信息，产品信息包括已购产品的交易用户的用户唯一标识；Obtain product information of products purchased within a specified time period, including the unique user ID of the transaction user who purchased the product;
根据用户唯一标识，查询出对应的用户信息；According to the user's unique identifier, query the corresponding user information;
根据用户信息和产品信息，生成交易数据。Generate transaction data based on user information and product information.
优选的，通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果，包括：Preferably, the transaction users are clustered according to the user information by using a clustering algorithm to obtain user clustering results, including:
根据各交易用户的交易特征，按照预设的聚类中心数量，从交易用户中选取出多个样本用户；According to the transaction characteristics of each transaction user, multiple sample users are selected from the transaction users according to the preset number of cluster centers;
通过聚类算法，将多个样本用户分别作为做个初始聚类中心对交易用户进行聚类，生成用户聚类结果，用户聚类结果包括多个用户类型和每个用户类型对应的交易用户。Through the clustering algorithm, multiple sample users are used as initial clustering centers to cluster transaction users and generate user clustering results. The user clustering results include multiple user types and transaction users corresponding to each user type.
优选的，根据各交易用户的交易特征，按照预设的聚类中心数量，从交易用户中选取出多个样本用户，包括：Preferably, according to the transaction characteristics of each transaction user and the preset number of cluster centers, a plurality of sample users are selected from the transaction users, including:
随机选取一个交易用户作为当前的样本用户，并按照预设的聚类特征获取当前的样本用户的聚类特征值；Randomly select a transaction user as the current sample user, and obtain the clustering feature value of the current sample user according to the preset clustering feature;
根据各交易用户的交易特征，生成每个交易用户与当前的样本用户之间的特征距离；According to the transaction characteristics of each transaction user, generate the characteristic distance between each transaction user and the current sample user;
从多个特征距离中选取出最大的特征距离对应的候选样本用户，并按照聚类特征获取候选样本用户的聚类特征值；Selecting a candidate sample user corresponding to the largest feature distance from multiple feature distances, and obtaining a clustering feature value of the candidate sample user according to the clustering feature;
判断候选样本用户的聚类特征值是否与当前的样本用户的聚类特征值相同；Determine whether the clustering feature value of the candidate sample user is the same as the clustering feature value of the current sample user;
若是，将候选样本用户对应的特征距离从生成的多个特征距离中滤除；If so, the feature distance corresponding to the candidate sample user is filtered out from the generated multiple feature distances;
从滤除后的多个特征距离中选取出最大的特征距离对应的候选样本用户，并继续执行按照聚类特征获取候选样本用户的聚类特征值的步骤；Selecting a candidate sample user corresponding to the largest characteristic distance from the filtered multiple characteristic distances, and continuing to perform the step of obtaining clustering characteristic values of the candidate sample users according to the clustering characteristics;
若否，将候选样本用户确定为当前的样本用户；If not, the candidate sample user is determined as the current sample user;
判断样本用户的数量是否小于聚类中心数量；Determine whether the number of sample users is less than the number of cluster centers;
若是，继续执行根据各交易用户的交易特征，生成每个交易用户与当前的样本用户之间的特征距离的步骤。If so, continue to execute the step of generating a characteristic distance between each transaction user and the current sample user according to the transaction characteristics of each transaction user.
优选的，产品信息包括产品收益率和用户信息包括用户收益率，用户聚类结果包括多个用户类型；Preferably, the product information includes product rate of return and the user information includes user rate of return, and the user clustering result includes multiple user types;
根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息，包括：Based on the user clustering results, product information, and user information of transaction users who have purchased products, generate pre-recommended product information corresponding to the user type, including:
对每个用户类型中交易用户的用户收益率进行排序，统计出指定排序的用户收益率对应的预推荐产品和预推荐产品用户群；Sort the user yields of the trading users in each user type, and calculate the pre-recommended products and pre-recommended product user groups corresponding to the user yields of the specified sort;
根据预推荐产品的产品收益率和预推荐产品用户群中的购买人数，生成预推荐产品的产品分数；Generate a product score for the pre-recommended product based on the product yield of the pre-recommended product and the number of purchasers in the pre-recommended product user group;
根据产品分数、产品信息和用户类型，生成用户类型对应的预推荐产品信息。Based on the product score, product information and user type, pre-recommended product information corresponding to the user type is generated.
优选的，在通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息之前，还包括：Preferably, before performing classification prediction on the pre-generated user portrait of the target user according to the user clustering result and the pre-recommended product information corresponding to the user type by using a classification algorithm and generating the target recommended product information, the method further includes:
响应于目标用户的产品申购登录请求，获取目标用户的登录信息；In response to a product purchase login request from a target user, obtaining login information of the target user;
将预先设置的产品推荐问答页面进行可视化展示，并接收目标用户输入的产品偏好信息；Visualize the pre-set product recommendation question and answer page and receive product preference information input by target users;
根据登录信息和产品偏好信息，生成目标用户的用户画像。Generate user profiles of target users based on login information and product preference information.
优选的，通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息，包括：Preferably, a classification algorithm is used to classify and predict the pre-generated user portrait of the target user according to the user clustering result and the pre-recommended product information corresponding to the user type, and generate the target recommended product information, including:
通过K最邻近算法，根据用户聚类结果，对用户画像进行分类，生成目标用户所属的目标类型；Through the K nearest neighbor algorithm, the user portraits are classified according to the user clustering results to generate the target type to which the target user belongs;
将目标类型与用户类型进行匹配，查询出对应的预推荐产品信息；Match the target type with the user type and query the corresponding pre-recommended product information;
根据用户画像，对预推荐产品信息进行筛选，生成目标推荐产品信息。According to the user portrait, the pre-recommended product information is filtered to generate the target recommended product information.
本发明还公开了一种产品推荐装置，包括：The present invention also discloses a product recommendation device, comprising:
交易数据获取单元，用于获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；A transaction data acquisition unit, used to acquire transaction data of purchased products, the transaction data including product information and user information of the transaction user of the purchased product;
用户聚类单元，用于通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；A user clustering unit, used to cluster transaction users according to user information through a clustering algorithm to obtain user clustering results;
预推荐产品信息单元，用于根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；A pre-recommended product information unit, used to generate pre-recommended product information corresponding to the user type according to the user clustering result, product information and user information of the transaction user who has purchased the product;
分类预测单元，用于通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息。The classification prediction unit is used to perform classification prediction on the pre-generated user portrait of the target user through a classification algorithm according to the user clustering results and the pre-recommended product information corresponding to the user type, and generate target recommended product information.
本发明还公开了一种计算机可读介质，其上存储有计算机程序，该程序被处理器执行时实现如上所述方法。The present invention also discloses a computer-readable medium on which a computer program is stored. When the program is executed by a processor, the method described above is implemented.
本发明还公开了一种计算机设备，包括存储器和处理器，所述存储器用于存储包括程序指令的信息，所述处理器用于控制程序指令的执行，所述处理器执行所述程序时实现如上所述方法。The present invention also discloses a computer device, including a memory and a processor, wherein the memory is used to store information including program instructions, the processor is used to control the execution of program instructions, and the processor implements the above method when executing the program.
本发明还公开了一种计算机程序产品，包括计算机程序/指令，计算机程序/指令被处理器执行时实现如上所述方法。The present invention also discloses a computer program product, including a computer program/instruction, and the method described above is implemented when the computer program/instruction is executed by a processor.
本发明获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息，对已购产品的交易数据对用户进行聚类分析，结合问卷应答以无侵入的方式嵌入原有的产品申购页面，通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，结合用户偏好和产品收益，更加精准的预测用户的购买产品，更加快捷的促成交易，提高产品推荐精度，从而提升用户体验，进而提升产品的销售效率。The present invention obtains transaction data of purchased products, where the transaction data includes product information and user information of transaction users of purchased products; clusters transaction users according to user information through a clustering algorithm to obtain user clustering results; generates pre-recommended product information corresponding to user types according to user clustering results, product information and user information of transaction users of purchased products; classifies and predicts user portraits of pre-generated target users according to user clustering results and pre-recommended product information corresponding to user types through a classification algorithm to generate target recommended product information, performs cluster analysis on users on transaction data of purchased products, embeds the data into an original product purchase page in a non-invasive manner in combination with questionnaire answers, accurately classifies users by improving user features and instantly updating user information, more accurately predicts users' purchase products in combination with user preferences and product benefits, facilitates transactions more quickly, improves product recommendation accuracy, thereby improving user experience and further improving product sales efficiency.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术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的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
图1为本发明实施例提供的一种产品推荐方法的流程图；FIG1 is a flow chart of a product recommendation method provided by an embodiment of the present invention;
图2为本发明实施例提供的又一种产品推荐方法的流程图；FIG2 is a flow chart of another product recommendation method provided by an embodiment of the present invention;
图3为本发明实施例提供的一种产品推荐装置的结构示意图；FIG3 is a schematic diagram of the structure of a product recommendation device provided by an embodiment of the present invention;
图4为本发明实施例提供的一种计算机设备的结构示意图。FIG. 4 is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention.
具体实施方式DETAILED DESCRIPTION
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，所描述的实施例仅仅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做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 following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
需要说明的是，本申请公开的一种产品推荐方法和装置可用于人工智能技术领域，也可用于除人工智能技术领域之外的任意领域，本申请公开的一种产品推荐方法和装置的应用领域不做限定。It should be noted that the product recommendation method and device disclosed in the present application can be used in the field of artificial intelligence technology, and can also be used in any field outside the field of artificial intelligence technology. The application field of the product recommendation method and device disclosed in the present application is not limited.
为了便于理解本申请提供的技术方案，下面先对本申请技术方案的相关内容进行说明。本发明整体由数据准备和产品预测两个部分组成，数据准备是分析历史数据给产品预测提供可参考的数据，产品预测是基于提供的数据对当前用户进行分类。数据准备对过去购买过理财产品的用户进行研究，分析用户的行为习惯，最后分类保存优质用户购买过的理财产品作为未来推荐的参考。产品预测面向使用“理财产品推荐助手”的用户，首先获取并补全用户的特征，之后参考与该用户相似的历史优质用户的购买产品对该用户进行产品推荐。In order to facilitate the understanding of the technical solution provided by the present application, the relevant contents of the technical solution of the present application are first described below. The present invention as a whole consists of two parts: data preparation and product prediction. Data preparation is to analyze historical data to provide reference data for product prediction, and product prediction is to classify current users based on the provided data. Data preparation studies users who have purchased financial products in the past, analyzes user behavior habits, and finally classifies and saves financial products purchased by high-quality users as a reference for future recommendations. Product prediction is for users who use the "Financial Product Recommendation Assistant". First, the user's features are obtained and completed, and then product recommendations are made to the user with reference to the purchased products of historical high-quality users similar to the user.
本发明使用聚类(K-means)算法，结合预先配置的“理财产品推荐助手”模块，对用户推荐理财产品流程和结果进行优化。理财产品推荐助手以小标签形式嵌入到理财产品申购页面，小标签固定在屏幕左侧下方。点击标签后在当前页面以弹窗的形式打开。在该弹窗中系统将会以问卷的形式收集用户的信息。此处认为已在平台处开通账户，平台已经获取用户年龄、性别、学历等基本信息。采集用户意向购买产品的风险等级、期限、用户本人的投资经验此类尚未获取到的信息。在采集完所有所需信息之后，结合用户的所有特征构建用户画像，并且根据用户画像进行归类，在弹窗中展示该类用户中推荐排名最高的产品。The present invention uses a clustering (K-means) algorithm, combined with a pre-configured "financial product recommendation assistant" module, to optimize the process and results of recommending financial products to users. The financial product recommendation assistant is embedded in the financial product subscription page in the form of a small label, and the small label is fixed at the bottom left of the screen. After clicking the label, it will be opened in the form of a pop-up window on the current page. In this pop-up window, the system will collect user information in the form of a questionnaire. It is assumed here that an account has been opened at the platform, and the platform has obtained basic information such as the user's age, gender, and education. Collect information that has not yet been obtained, such as the risk level, term, and investment experience of the user's intended product. After collecting all the required information, a user portrait is constructed in combination with all the user's characteristics, and classified according to the user portrait, and the highest recommended products among this type of users are displayed in the pop-up window.
理财产品推荐助手是以无侵入的方式嵌入原有的理财申购页面，模块独立于原有其他功能，不会对存量功能造成影响。而且给了用户一种新的方式来选择意向购买的产品，该方式通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，更加精准的预测用户的购买产品，更加快捷的促成交易。The financial product recommendation assistant is embedded in the original financial subscription page in a non-invasive way. The module is independent of other original functions and will not affect the existing functions. It also gives users a new way to choose the products they intend to buy. This method accurately classifies users by improving user characteristics and updating user information in real time, more accurately predicts users' purchase products, and facilitates transactions more quickly.
下面以产品推荐装置作为执行主体为例，说明本发明实施例提供的产品推荐方法的实现过程。可理解的是，本发明实施例提供的产品推荐方法的执行主体包括但不限于产品推荐装置。The following takes the product recommendation device as an example to illustrate the implementation process of the product recommendation method provided by the embodiment of the present invention. It is understandable that the execution subject of the product recommendation method provided by the embodiment of the present invention includes but is not limited to the product recommendation device.
图1为本发明实施例提供的一种产品推荐方法的流程图，如图1所示，该方法包括：FIG1 is a flow chart of a product recommendation method provided by an embodiment of the present invention. As shown in FIG1 , the method includes:
步骤101、获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息。Step 101: Acquire transaction data of purchased products, where the transaction data includes product information and user information of the transaction user of the purchased product.
步骤102、通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果。Step 102: Cluster the transaction users according to the user information through a clustering algorithm to obtain a user clustering result.
步骤103、根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息。Step 103: Generate pre-recommended product information corresponding to the user type based on the user clustering result, product information, and user information of the transaction user who has purchased the product.
步骤104、通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息。Step 104: Using a classification algorithm, based on the user clustering results and the pre-recommended product information corresponding to the user type, a classification prediction is performed on the pre-generated user portrait of the target user to generate target recommended product information.
本发明实施例提供的技术方案中，获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息，对已购产品的交易数据对用户进行聚类分析，结合问卷应答以无侵入的方式嵌入原有的产品申购页面，通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，结合用户偏好和产品收益，更加精准的预测用户的购买产品，更加快捷的促成交易，提高产品推荐精度，从而提升用户体验，进而提升产品的销售效率。In the technical solution provided by the embodiment of the present invention, transaction data of purchased products are obtained, and the transaction data includes product information and user information of transaction users of purchased products; transaction users are clustered according to user information through a clustering algorithm to obtain user clustering results; pre-recommended product information corresponding to the user type is generated according to the user clustering results, product information and user information of transaction users of purchased products; user portraits of pre-generated target users are classified and predicted according to the user clustering results and the pre-recommended product information corresponding to the user type through a classification algorithm, target recommended product information is generated, and cluster analysis is performed on the transaction data of purchased products, and the data is embedded in the original product purchase page in a non-invasive manner in combination with questionnaire answers, and users are accurately classified by improving user features and instantly updating user information, and users' purchase products are more accurately predicted in combination with user preferences and product benefits, transactions are facilitated more quickly, and the accuracy of product recommendations is improved, thereby improving user experience and further improving product sales efficiency.
图2为本发明实施例提供的又一种产品推荐方法的流程图，如图2所示，该方法包括：FIG. 2 is a flow chart of another product recommendation method provided by an embodiment of the present invention. As shown in FIG. 2 , the method includes:
步骤201、获取指定时间段内已购产品的产品信息。Step 201: Obtain product information of products purchased within a specified time period.
本发明实施例中，各步骤由产品推荐装置执行。In the embodiment of the present invention, each step is performed by a product recommendation device.
本发明实施例中，指定时间段可以根据实际需求进行设置，本发明实施例对此不做限定。作为一种可选方案，指定时间段为距离当前时间之前的6个月。In the embodiment of the present invention, the specified time period can be set according to actual needs, and the embodiment of the present invention does not limit this. As an optional solution, the specified time period is 6 months before the current time.
本发明实施例中，从数据库中的购买记录中获取已购产品的产品信息，产品信息包括但不限于购买该产品的交易用户的用户唯一标识、产品收益率、产品唯一标识、产品币种、所属公司、所属模块、风险等级、现值、到期期限。其中，用户唯一标识用于标识唯一用户，产品唯一标识用于标识唯一产品。In an embodiment of the present invention, product information of a purchased product is obtained from the purchase record in the database, and the product information includes but is not limited to the user unique identifier of the transaction user who purchased the product, the product yield, the product unique identifier, the product currency, the company to which it belongs, the module to which it belongs, the risk level, the present value, and the expiration date. Among them, the user unique identifier is used to identify a unique user, and the product unique identifier is used to identify a unique product.
进一步地，为了提高后续计算效率，对产品信息进行提取。作为一种可选方案，从产品信息中提取出产品收益率排名前五的产品对应的产品信息，可以降低数据量，有效提升后续计算效率。Furthermore, in order to improve the subsequent calculation efficiency, the product information is extracted. As an optional solution, the product information corresponding to the top five products in terms of product yield is extracted from the product information, which can reduce the amount of data and effectively improve the subsequent calculation efficiency.
步骤202、根据用户唯一标识，查询出对应的用户信息。Step 202: Query the corresponding user information according to the user unique identifier.
本发明实施例中，数据库中存储有已购产品的交易用户的基本信息，通过产品信息中的用户唯一标识与存储的基本信息中的用户唯一标识进行匹配，查询出对应的基本信息；将查询出的基本信息确定为交易用户的用户信息。用户信息包括但不限于年龄、性别、学历、收入、消费习惯、资产金额、负债金额、用户收益率。In the embodiment of the present invention, the database stores basic information of transaction users who have purchased products, and the corresponding basic information is queried by matching the user unique identifier in the product information with the user unique identifier in the stored basic information; the queried basic information is determined as the user information of the transaction user. The user information includes but is not limited to age, gender, education, income, consumption habits, asset amount, liability amount, and user rate of return.
步骤203、根据用户信息和产品信息，生成交易数据。Step 203: Generate transaction data based on user information and product information.
具体地，按照用户唯一标识，将产品信息和用户信息进行汇总，生成交易数据。Specifically, product information and user information are aggregated according to the user's unique identifier to generate transaction data.
步骤204、根据各交易用户的交易特征，按照预设的聚类中心数量，从交易用户中选取出多个样本用户。Step 204: Select multiple sample users from the transaction users according to the transaction characteristics of each transaction user and the preset number of cluster centers.
本发明实施例中，聚类中心数量可以根据实际需求进行设置，本发明实施例对此不做限定。In the embodiment of the present invention, the number of cluster centers can be set according to actual needs, and the embodiment of the present invention does not limit this.
本发明实施例中，使用K-means聚类算法进行用户聚类，需要选取若干个对象作为初始的聚类中心。步骤204具体包括：In the embodiment of the present invention, the K-means clustering algorithm is used to cluster users, and several objects need to be selected as initial cluster centers. Step 204 specifically includes:
步骤2041、随机选取一个交易用户作为当前的样本用户，并按照预设的聚类特征获取当前的样本用户的聚类特征值。Step 2041: randomly select a transaction user as the current sample user, and obtain the clustering feature value of the current sample user according to the preset clustering feature.
本发明实施例中，聚类特征可以根据实际需求进行设置，本发明实施例对此不做限定。作为一种可选方案，本发明实施例选取的聚类特征为14-35岁、35-60岁、60岁及以上三个不同年龄区间以及性别，也即：聚类中心数量为6。In the embodiment of the present invention, the clustering features can be set according to actual needs, and the embodiment of the present invention does not limit this. As an optional solution, the clustering features selected by the embodiment of the present invention are three different age ranges of 14-35 years old, 35-60 years old, and 60 years old and above, and gender, that is, the number of cluster centers is 6.
值得说明的是，聚类特征支持银行自行定义，除了年龄和性别之外还可以选择其他不同维度的特征对已购产品的交易用户进行聚类，如不同资产等级、信用等级。It is worth noting that banks can define clustering features by themselves. In addition to age and gender, they can also select features of other dimensions to cluster transaction users who have purchased products, such as different asset levels and credit levels.
步骤2042、根据各交易用户的交易特征，生成每个交易用户与当前的样本用户之间的特征距离。Step 2042: Generate a characteristic distance between each transaction user and the current sample user based on the transaction characteristics of each transaction user.
本发明实施例中，从交易数据中筛选出数值型特征，得到交易特征，例如：资产金额和负债金额。In the embodiment of the present invention, numerical features are screened out from the transaction data to obtain transaction features, such as asset amounts and liability amounts.
具体地，通过对各交易用户的交易特征和当前的样本用户的交易特征进行欧氏距离计算，生成每个交易用户与当前的样本用户之间的特征距离。其中，d(xi,y)为交易用户xi与当前的样本用户y之间的特征距离，xi1……xim为交易用户xi的交易特征，y1……ym为当前的样本用户y的交易特征。Specifically, through The Euclidean distance between the transaction characteristics of each transaction user and the current sample user is calculated to generate the characteristic distance between each transaction user and the current sample user. Among them, d(xi ,y) is the characteristic distance between transactionuserxi and the current sample usery,xi1 ...xim is the transaction characteristics of transactionuserxi ,andy1 ...ym is the transaction characteristics of the current sample usery.
步骤2043、从多个特征距离中选取出最大的特征距离对应的候选样本用户，并按照聚类特征获取候选样本用户的聚类特征值。Step 2043: Select a candidate sample user corresponding to the largest characteristic distance from multiple characteristic distances, and obtain a clustering characteristic value of the candidate sample user according to the clustering characteristic.
本发明实施例中，对多个特征距离进行比较，选取出最大的特征距离；将最大的特征距离对应的交易用户确定为候选样本用户；获取候选样本用户的聚类特征值。In the embodiment of the present invention, multiple feature distances are compared to select the largest feature distance; the transaction user corresponding to the largest feature distance is determined as the candidate sample user; and the clustering feature value of the candidate sample user is obtained.
例如：当前的样本用户的特征值为20岁、女性，符合14-35岁年龄区间的女性的聚类特征值；候选样本用户的特征值为30岁、男性，符合14-35岁年龄区间的男性的聚类特征值。For example, the characteristic value of the current sample user is 20 years old and female, which meets the cluster characteristic value of females in the age range of 14-35 years old; the characteristic value of the candidate sample user is 30 years old and male, which meets the cluster characteristic value of males in the age range of 14-35 years old.
步骤2044、判断候选样本用户的聚类特征值是否与当前的样本用户的聚类特征值相同，若是，执行步骤2045；若否，执行步骤2047。Step 2044 , determine whether the clustering feature value of the candidate sample user is the same as the clustering feature value of the current sample user. If so, execute step 2045 ; if not, execute step 2047 .
本发明实施例中，若候选样本用户的聚类特征值与当前的样本用户的聚类特征值相同，表明候选样本用户和当前的样本用户属于同一个用户类型，继续执行步骤2045；若候选样本用户的聚类特征值与当前的样本用户的聚类特征值不同，表明候选样本用户和当前的样本用户不属于同一个用户类型，继续执行步骤2047。In this embodiment of the present invention, if the clustering feature value of the candidate sample user is the same as the clustering feature value of the current sample user, it indicates that the candidate sample user and the current sample user belong to the same user type, and step 2045 is continued to be executed; if the clustering feature value of the candidate sample user is different from the clustering feature value of the current sample user, it indicates that the candidate sample user and the current sample user do not belong to the same user type, and step 2047 is continued to be executed.
例如：当前的样本用户是20岁、女性，其聚类特征值为14-35岁年龄区间的女性；候选样本用户是30岁、男性，其聚类特征值为14-35岁年龄区间的男性，则候选样本用户的聚类特征值与当前的样本用户的聚类特征值不同，表明候选样本用户和当前的样本用户不属于同一个用户类型，需要重新选择样本用户，继续执行步骤2047。For example: the current sample user is 20 years old, female, and her clustering feature value is female in the age range of 14-35 years old; the candidate sample user is 30 years old, male, and her clustering feature value is male in the age range of 14-35 years old. The clustering feature value of the candidate sample user is different from the clustering feature value of the current sample user, indicating that the candidate sample user and the current sample user do not belong to the same user type, and it is necessary to reselect the sample user and continue to execute step 2047.
例如：当前的样本用户是20岁、女性，其聚类特征值为14-35岁年龄区间的女性；候选样本用户是30岁、女性，其聚类特征值为14-35岁年龄区间的女性，则候选样本用户的聚类特征值与当前的样本用户的聚类特征值相同，表明候选样本用户和当前的样本用户属于同一个用户类型，继续执行步骤2045。For example: the current sample user is 20 years old, female, and her clustering feature value is female in the age range of 14-35 years old; the candidate sample user is 30 years old, female, and her clustering feature value is female in the age range of 14-35 years old, then the clustering feature value of the candidate sample user is the same as the clustering feature value of the current sample user, indicating that the candidate sample user and the current sample user belong to the same user type, and continue to execute step 2045.
步骤2045、将候选样本用户对应的特征距离从生成的多个特征距离中滤除。Step 2045: Filter out the feature distances corresponding to the candidate sample users from the generated multiple feature distances.
本发明实施例中，若候选样本用户和当前的样本用户不属于同一个用户类型，需要重新选取下一个样本用户，则将候选样本用户对应的特征距离从生成的多个特征距离中滤除，避免选取下一个样本用户时重复选取到被排除的用户，节约计算资源。In an embodiment of the present invention, if the candidate sample user and the current sample user do not belong to the same user type and the next sample user needs to be reselected, the feature distance corresponding to the candidate sample user is filtered out from the generated multiple feature distances to avoid repeated selection of the excluded user when selecting the next sample user, thereby saving computing resources.
步骤2046、从滤除后的多个特征距离中选取出最大的特征距离对应的候选样本用户，并继续执行步骤2043中的按照聚类特征获取候选样本用户的聚类特征值。Step 2046: Select the candidate sample user corresponding to the largest feature distance from the filtered multiple feature distances, and continue to execute step 2043 to obtain the cluster feature value of the candidate sample user according to the cluster feature.
本发明实施例中，对滤除后的多个特征距离进行比较，选取出最大的特征距离；将最大的特征距离对应的交易用户确定为候选样本用户；获取候选样本用户的聚类特征值与当前的样本用户的聚类特征值进行判断，直至选取到与当前的样本用户不属于同一个用户类型的用户。In an embodiment of the present invention, multiple feature distances after filtering are compared to select the largest feature distance; the transaction user corresponding to the largest feature distance is determined as a candidate sample user; the cluster feature value of the candidate sample user is obtained and judged with the cluster feature value of the current sample user until a user who does not belong to the same user type as the current sample user is selected.
步骤2047、将候选样本用户确定为当前的样本用户。Step 2047: Determine the candidate sample user as the current sample user.
本发明实施例中，若候选样本用户和当前的样本用户不属于同一个用户类型，存储当前的样本用户，并将候选样本用户更新为当前的样本用户，在下一个样本用户的选取中，以当前的样本用户为基准进行特征距离比较。In the embodiment of the present invention, if the candidate sample user and the current sample user do not belong to the same user type, the current sample user is stored, and the candidate sample user is updated to the current sample user. In the selection of the next sample user, the feature distance comparison is performed based on the current sample user.
例如：当前的样本用户为样本1，候选样本用户为样本2，若样本2和样本1不属于同一个用户类型，存储样本1，并将样本2更新为当前的样本用户，在下一个样本用户的选取中，以样本2为基准进行特征距离比较。For example: the current sample user is sample 1, and the candidate sample user is sample 2. If sample 2 and sample 1 do not belong to the same user type, sample 1 is stored and sample 2 is updated as the current sample user. In the selection of the next sample user, sample 2 is used as the benchmark for feature distance comparison.
步骤2048、判断样本用户的数量是否小于聚类中心数量，若是，执行步骤2042；若否，执行步骤2049。Step 2048, determine whether the number of sample users is less than the number of cluster centers, if so, execute step 2042; if not, execute step 2049.
本发明实施例中，若已存储的样本用户的数量小于聚类中心数量，表明还未达到初始聚类中心的数量要求，继续执行步骤2042；若已存储的样本用户的数量大于或等于聚类中心数量，表明已达到初始聚类中心的数量要求，继续执行步骤2049。In an embodiment of the present invention, if the number of stored sample users is less than the number of cluster centers, it indicates that the number requirement of the initial cluster centers has not been met, and step 2042 is continued; if the number of stored sample users is greater than or equal to the number of cluster centers, it indicates that the number requirement of the initial cluster centers has been met, and step 2049 is continued.
步骤2049、确定出多个样本用户。Step 2049: Determine multiple sample users.
本发明实施例中，确定出的多个样本用户包括已存储的样本用户和当前的样本用户，例如：已存储的样本1和当前的样本用户样本2。In the embodiment of the present invention, the determined multiple sample users include stored sample users and current sample users, for example: stored sample 1 and current sample user sample 2.
步骤205、通过聚类算法，将多个样本用户分别作为做个初始聚类中心对交易用户进行聚类，生成用户聚类结果。Step 205: cluster the transaction users using a clustering algorithm by taking a plurality of sample users as initial clustering centers to generate user clustering results.
本发明实施例中，聚类算法为K-means算法。具体地，通过K-means算法，将多个样本用户分别作为做个初始聚类中心对交易用户进行聚类，生成用户聚类结果。In the embodiment of the present invention, the clustering algorithm is a K-means algorithm. Specifically, the K-means algorithm is used to cluster the transaction users using multiple sample users as initial clustering centers to generate user clustering results.
本发明实施例中，用户聚类结果包括多个用户类型和每个用户类型对应的交易用户。例如：根据性别和年龄将交易用户聚类为六类不同的用户类型，分别是年轻男性、年轻女性、中年男性、中年女性、老年男性、老年女性。In the embodiment of the present invention, the user clustering result includes multiple user types and transaction users corresponding to each user type. For example, transaction users are clustered into six different user types according to gender and age, namely young men, young women, middle-aged men, middle-aged women, old men, and old women.
本发明实施例中，经过聚类之后银行可以更加快速的定位用户的类别，制定相应的营销方案。In the embodiment of the present invention, after clustering, the bank can locate the category of users more quickly and formulate corresponding marketing plans.
步骤206、对每个用户类型中交易用户的用户收益率进行排序，统计出指定排序的用户收益率对应的预推荐产品和预推荐产品用户群。Step 206: sort the user yields of the transaction users in each user type, and calculate the pre-recommended products and pre-recommended product user groups corresponding to the user yields of the specified sort.
具体地，对每个用户类型中的交易用户按照用户收益率进行降序排序，选取指定排序的用户收益率的交易用户；将选取出的用户确定为预推荐产品用户群；获取预推荐产品用户群中每个交易用户的已购产品，将获取的已购产品确定为预推荐产品。其中，预推荐产品用户群和预推荐产品与用户类型相对应。Specifically, the transaction users in each user type are sorted in descending order according to the user rate of return, and the transaction users with the specified sorted user rate of return are selected; the selected users are determined as the pre-recommended product user group; the purchased products of each transaction user in the pre-recommended product user group are obtained, and the obtained purchased products are determined as pre-recommended products. The pre-recommended product user group and the pre-recommended product correspond to the user type.
本发明实施例中，指定排序可以根据实际需求进行设置，本发明实施例对此不做限定。作为一种可选方案，指定排序为用户收益率排名前10％。In the embodiment of the present invention, the specified ranking can be set according to actual needs, and the embodiment of the present invention does not limit this. As an optional solution, the specified ranking is the top 10% of the user yield ranking.
步骤207、根据预推荐产品的产品收益率和预推荐产品用户群中的购买人数，生成预推荐产品的产品分数。Step 207: Generate a product score for the pre-recommended product based on the product rate of return of the pre-recommended product and the number of purchasers in the user group of the pre-recommended product.
本发明实施例中，查询出预推荐产品用户群中购买该预推荐产品的购买人数；将预推荐产品的产品收益率与该预推荐产品对应的购买人数相乘，生成该预推荐产品的产品分数。In the embodiment of the present invention, the number of purchasers of the pre-recommended product in the pre-recommended product user group is queried; the product yield of the pre-recommended product is multiplied by the number of purchasers corresponding to the pre-recommended product to generate a product score for the pre-recommended product.
步骤208、根据产品分数、产品信息和用户类型，生成用户类型对应的预推荐产品信息。Step 208: Generate pre-recommended product information corresponding to the user type based on the product score, product information and user type.
本发明实施例中，根据用户类型对应的多个预推荐产品的产品唯一标识，查询出每个预推荐产品对应的产品分数和产品信息；根据每个预推荐产品对应的产品分数和产品信息，生成该用户类型对应的预推荐产品信息。预推荐产品信息包括但不限于产品唯一标识、产品分数、产品收益率、产品唯一标识、产品币种、所属公司、所属模块、风险等级、现值、到期期限、用户类型。In the embodiment of the present invention, the product score and product information corresponding to each pre-recommended product are queried according to the product unique identifiers of multiple pre-recommended products corresponding to the user type; and the pre-recommended product information corresponding to the user type is generated according to the product score and product information corresponding to each pre-recommended product. The pre-recommended product information includes but is not limited to the product unique identifier, product score, product yield, product unique identifier, product currency, company, module, risk level, present value, maturity, and user type.
进一步地，将预推荐产品信息保存至数据库中，所有非金额的项目都以枚举值的形式保存。Furthermore, the pre-recommended product information is saved in the database, and all non-monetary items are saved in the form of enumeration values.
本发明实施例中，预推荐产品信息是优质用户购买过的理财产品，具有高优先级，可以作为后续推荐产品的参考。In the embodiment of the present invention, the pre-recommended product information is the financial products purchased by high-quality users, has a high priority, and can be used as a reference for subsequent recommended products.
步骤209、响应于目标用户的产品申购登录请求，获取目标用户的登录信息。Step 209: In response to the product purchase login request of the target user, the login information of the target user is obtained.
本发明实施例中，目标用户为需要推荐产品的用户。用户打开应用程序，进入理财产品申购页面进行登录，此时用户已经在网站或应用程序中注册过账号，能够获取到该用户的登录信息，登录信息包括但不限于年龄、性别、学历、收入、消费习惯、资产金额、负债金额、用户收益率。In the embodiment of the present invention, the target user is a user who needs to recommend products. The user opens the application and enters the financial product subscription page to log in. At this time, the user has registered an account on the website or application and can obtain the user's login information, which includes but is not limited to age, gender, education, income, consumption habits, asset amount, liability amount, and user rate of return.
步骤210、将预先设置的产品推荐问答页面进行可视化展示，并接收目标用户输入的产品偏好信息。Step 210: Visually display the pre-set product recommendation question and answer page, and receive product preference information input by the target user.
本发明实施例中，用户登录成功后，理财产品推荐助手模块弹出理财产品推荐弹窗，弹窗中显示预先设置的产品推荐问答页面，产品推荐问答页面包括多个预先设置的问题，以供目标用户输入对应的回答；将目标用户输入的回答确定为产品偏好信息。In an embodiment of the present invention, after the user successfully logs in, the financial product recommendation assistant module pops up a financial product recommendation pop-up window, which displays a pre-set product recommendation question and answer page. The product recommendation question and answer page includes multiple pre-set questions for the target user to enter corresponding answers; the answers entered by the target user are determined as product preference information.
本发明实施例中，产品推荐问答页面中的问题包括但不限于：“您意向购买的理财产品的币种”、“您意向购买的理财产品的类别”、“您意向购买的理财产品的风险等级”、“您意向购买理财产品的板块”、“您的投资经验”，并以单选的形式由目标用户回答，答案以枚举值的形式保存。In an embodiment of the present invention, the questions in the product recommendation question and answer page include but are not limited to: "The currency of the financial product you intend to purchase", "The category of the financial product you intend to purchase", "The risk level of the financial product you intend to purchase", "The sector of the financial product you intend to purchase", and "Your investment experience", and are answered by the target user in the form of single choices, and the answers are saved in the form of enumerated values.
步骤211、根据登录信息和产品偏好信息，生成目标用户的用户画像。Step 211: Generate a user profile of the target user based on the login information and product preference information.
本发明实施例中，将目标用户的登录信息和产品偏好信息进行汇总，生成目标用户的用户画像。In the embodiment of the present invention, the login information and product preference information of the target user are summarized to generate a user profile of the target user.
步骤212、通过K最邻近算法，根据用户聚类结果，对用户画像进行分类，生成目标用户所属的目标类型。Step 212: Classify the user portraits according to the user clustering results through the K nearest neighbor algorithm to generate the target type to which the target user belongs.
具体地，将用户聚类结果和用户画像输入K最邻近算法，依据用户类型对该目标用户进行匹配分类，生成目标用户所属的目标类型。Specifically, the user clustering results and user portraits are input into the K nearest neighbor algorithm, the target user is matched and classified according to the user type, and the target type to which the target user belongs is generated.
步骤213、将目标类型与用户类型进行匹配，查询出对应的预推荐产品信息。Step 213: Match the target type with the user type and query the corresponding pre-recommended product information.
本发明实施例中，预推荐产品信息包括但不限于产品唯一标识、产品分数、产品收益率、产品唯一标识、产品币种、所属公司、所属模块、风险等级、现值、到期期限、用户类型。In an embodiment of the present invention, the pre-recommended product information includes but is not limited to the product unique identifier, product score, product yield, product unique identifier, product currency, company, module, risk level, present value, maturity, and user type.
步骤214、根据用户画像，对预推荐产品信息进行筛选，生成目标推荐产品信息。Step 214: Filter the pre-recommended product information according to the user portrait to generate target recommended product information.
作为一种可选方案，对预推荐产品信息按照预推荐产品信息中的产品分数进行降序排序，生成排序后的预推荐产品；选取指定分数排序的预推荐产品；将选取出的预推荐产品信息确定为目标推荐产品信息。其中，指定分数排序可以根据实际需求进行设置，本发明实施例对此不做限定。例如：指定分数排序为产品分数排名前五。As an optional solution, the pre-recommended product information is sorted in descending order according to the product scores in the pre-recommended product information to generate sorted pre-recommended products; the pre-recommended products sorted by the specified scores are selected; and the selected pre-recommended product information is determined as the target recommended product information. The specified score sorting can be set according to actual needs, and the embodiment of the present invention does not limit this. For example: the specified score sorting is the top five product scores.
进一步地，根据目标用户的产品偏好信息，从预推荐产品信息中选取出符合用户需求的产品。Furthermore, based on the product preference information of the target user, products that meet the user's needs are selected from the pre-recommended product information.
进一步地，将目标推荐产品信息进行可视化展示，以供用户选购所需的目标产品。Furthermore, the target recommended product information is displayed visually so that users can choose the desired target products.
本发明为有意愿购买理财产品的用户提供了一种挑选符合自身需求产品的新方案。用户可以结合自身的需求，从过去获得过高收益的用户所选择的产品中更加清晰的选择产品。银行可以结合自身原有用户的特征，对新用户可能购买的产品进行预测，进行更加精准的营销，实现理财产品销售效率和销量的提升。The present invention provides a new solution for users who are willing to purchase financial products to select products that meet their needs. Users can choose products more clearly from products selected by users who have obtained high returns in the past based on their own needs. Banks can predict the products that new users may purchase based on the characteristics of their original users, conduct more accurate marketing, and achieve improved sales efficiency and sales of financial products.
值得说明的是，本申请中技术方案中对数据的获取、存储、使用、处理等均符合法律法规的相关规定。本申请实施例中的用户信息均是通过合法合规途径获得，并且对用户信息的获取、存储、使用、处理等经过用户授权同意的。It is worth noting that the acquisition, storage, use, and processing of data in the technical solution of this application are in compliance with the relevant provisions of laws and regulations. The user information in the embodiments of this application is obtained through legal and compliant channels, and the acquisition, storage, use, and processing of user information are authorized and agreed by the user.
值得说明的是，本申请中采集的信息是经用户授权或者经过各方充分授权的信息和数据，并且相关数据的收集、存储、使用、加工、传输、提供、公开和应用等处理，均遵守相关国家和地区的相关法律法规和标准，采取了必要保密措施，不违背公序良俗，并提供有相应的操作入口，供用户选择授权或者拒绝。It is worth noting that the information collected in this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users to choose to authorize or refuse.
值得说明的是，本申请提供的技术方案，为用户提供相应的操作入口，供用户选择同意或者拒绝自动化决策结果；若用户选择拒绝，则进入专家决策流程。It is worth noting that the technical solution provided in this application provides users with corresponding operation entrances for them to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.
本发明实施例提供的产品推荐方法的技术方案中，获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息，对已购产品的交易数据对用户进行聚类分析，结合问卷应答以无侵入的方式嵌入原有的产品申购页面，通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，结合用户偏好和产品收益，更加精准的预测用户的购买产品，更加快捷的促成交易，提高产品推荐精度，从而提升用户体验，进而提升产品的销售效率。In the technical solution of the product recommendation method provided by the embodiment of the present invention, the transaction data of purchased products are obtained, and the transaction data includes product information and user information of transaction users of purchased products; through a clustering algorithm, transaction users are clustered according to user information to obtain user clustering results; based on the user clustering results, product information and user information of transaction users of purchased products, pre-recommended product information corresponding to the user type is generated; through a classification algorithm, based on the user clustering results and the pre-recommended product information corresponding to the user type, the user portrait of the pre-generated target user is classified and predicted to generate target recommended product information, the transaction data of purchased products are clustered and analyzed by users, and the original product purchase page is embedded in a non-invasive manner in combination with questionnaire answers, and users are accurately classified by improving user characteristics and updating user information in real time, and combined with user preferences and product benefits, the user's purchase products are predicted more accurately, and transactions are facilitated more quickly, thereby improving the accuracy of product recommendations, thereby improving user experience, and further improving product sales efficiency.
图3为本发明实施例提供的一种产品推荐装置的结构示意图，该装置用于执行上述产品推荐方法，如图3所示，该装置包括：交易数据获取单元11、用户聚类单元12、预推荐产品信息单元13和分类预测单元14。Figure 3 is a structural schematic diagram of a product recommendation device provided by an embodiment of the present invention. The device is used to execute the above-mentioned product recommendation method. As shown in Figure 3, the device includes: a transaction data acquisition unit 11, a user clustering unit 12, a pre-recommended product information unit 13 and a classification prediction unit 14.
交易数据获取单元11用于获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息。The transaction data acquisition unit 11 is used to acquire transaction data of purchased products, where the transaction data includes product information and user information of the transaction user of the purchased product.
用户聚类单元12用于通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果。The user clustering unit 12 is used to cluster transaction users according to user information through a clustering algorithm to obtain a user clustering result.
预推荐产品信息单元13用于根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息。The pre-recommended product information unit 13 is used to generate pre-recommended product information corresponding to the user type according to the user clustering result, product information and user information of the transaction user who has purchased the product.
分类预测单元14用于通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息。The classification prediction unit 14 is used to perform classification prediction on the pre-generated user portrait of the target user through a classification algorithm according to the user clustering result and the pre-recommended product information corresponding to the user type, and generate target recommended product information.
本发明实施例中，交易数据获取单元11具体用于获取指定时间段内已购产品的产品信息，产品信息包括已购产品的交易用户的用户唯一标识；根据用户唯一标识，查询出对应的用户信息；根据用户信息和产品信息，生成交易数据。In an embodiment of the present invention, the transaction data acquisition unit 11 is specifically used to obtain product information of products purchased within a specified time period, the product information including a unique user identifier of a transaction user who has purchased the product; query the corresponding user information based on the unique user identifier; and generate transaction data based on the user information and product information.
本发明实施例中，用户聚类单元12具体用于根据各交易用户的交易特征，按照预设的聚类中心数量，从交易用户中选取出多个样本用户；通过聚类算法，将多个样本用户分别作为做个初始聚类中心对交易用户进行聚类，生成用户聚类结果，用户聚类结果包括多个用户类型和每个用户类型对应的交易用户。In the embodiment of the present invention, the user clustering unit 12 is specifically used to select multiple sample users from the transaction users according to the transaction characteristics of each transaction user and the preset number of clustering centers; through the clustering algorithm, the multiple sample users are respectively used as initial clustering centers to cluster the transaction users to generate user clustering results, and the user clustering results include multiple user types and transaction users corresponding to each user type.
本发明实施例中，用户聚类单元12具体用于随机选取一个交易用户作为当前的样本用户，并按照预设的聚类特征获取当前的样本用户的聚类特征值；根据各交易用户的交易特征，生成每个交易用户与当前的样本用户之间的特征距离；从多个特征距离中选取出最大的特征距离对应的候选样本用户，并按照聚类特征获取候选样本用户的聚类特征值；判断候选样本用户的聚类特征值是否与当前的样本用户的聚类特征值相同；若是，将候选样本用户对应的特征距离从生成的多个特征距离中滤除；从滤除后的多个特征距离中选取出最大的特征距离对应的候选样本用户，并继续执行按照聚类特征获取候选样本用户的聚类特征值的步骤；若否，将候选样本用户确定为当前的样本用户；判断样本用户的数量是否小于聚类中心数量；若是，继续执行根据各交易用户的交易特征，生成每个交易用户与当前的样本用户之间的特征距离的步骤。In the embodiment of the present invention, the user clustering unit 12 is specifically used to randomly select a transaction user as the current sample user, and obtain the clustering feature value of the current sample user according to the preset clustering feature; generate the feature distance between each transaction user and the current sample user according to the transaction feature of each transaction user; select the candidate sample user corresponding to the largest feature distance from multiple feature distances, and obtain the clustering feature value of the candidate sample user according to the clustering feature; judge whether the clustering feature value of the candidate sample user is the same as the clustering feature value of the current sample user; if so, filter the feature distance corresponding to the candidate sample user from the generated multiple feature distances; select the candidate sample user corresponding to the largest feature distance from the filtered multiple feature distances, and continue to execute the step of obtaining the clustering feature value of the candidate sample user according to the clustering feature; if not, determine the candidate sample user as the current sample user; judge whether the number of sample users is less than the number of cluster centers; if so, continue to execute the step of generating the feature distance between each transaction user and the current sample user according to the transaction features of each transaction user.
本发明实施例中，产品信息包括产品收益率和用户信息包括用户收益率，用户聚类结果包括多个用户类型；预推荐产品信息单元13具体用于对每个用户类型中交易用户的用户收益率进行排序，统计出指定排序的用户收益率对应的预推荐产品和预推荐产品用户群；根据预推荐产品的产品收益率和预推荐产品用户群中的购买人数，生成预推荐产品的产品分数；根据产品分数、产品信息和用户类型，生成用户类型对应的预推荐产品信息。In an embodiment of the present invention, product information includes product yield rate and user information includes user yield rate, and the user clustering result includes multiple user types; the pre-recommended product information unit 13 is specifically used to sort the user yield rates of trading users in each user type, and count the pre-recommended products and pre-recommended product user groups corresponding to the user yield rates of the specified sort; generate a product score for the pre-recommended product based on the product yield rate of the pre-recommended product and the number of purchasers in the pre-recommended product user group; generate pre-recommended product information corresponding to the user type based on the product score, product information and user type.
本发明实施例中，该装置还包括：登录信息获取单元15、可视化单元16和用户画像生成单元17。In the embodiment of the present invention, the device further includes: a login information acquisition unit 15 , a visualization unit 16 and a user portrait generation unit 17 .
登录信息获取单元15用于响应于目标用户的产品申购登录请求，获取目标用户的登录信息。The login information acquisition unit 15 is used to acquire the login information of the target user in response to the product purchase login request of the target user.
可视化单元16用于将预先设置的产品推荐问答页面进行可视化展示，并接收目标用户输入的产品偏好信息。The visualization unit 16 is used to visualize the preset product recommendation question and answer page and receive product preference information input by the target user.
用户画像生成单元17用于根据登录信息和产品偏好信息，生成目标用户的用户画像。The user portrait generating unit 17 is used to generate a user portrait of the target user according to the login information and the product preference information.
本发明实施例中，分类预测单元14具体用于通过K最邻近算法，根据用户聚类结果，对用户画像进行分类，生成目标用户所属的目标类型；将目标类型与用户类型进行匹配，查询出对应的预推荐产品信息；根据用户画像，对预推荐产品信息进行筛选，生成目标推荐产品信息。In an embodiment of the present invention, the classification prediction unit 14 is specifically used to classify user portraits according to user clustering results through the K nearest neighbor algorithm, and generate a target type to which the target user belongs; match the target type with the user type, and query the corresponding pre-recommended product information; according to the user portrait, filter the pre-recommended product information to generate target recommended product information.
本发明实施例的方案中，获取已购产品的交易数据，交易数据包括产品信息和已购产品的交易用户的用户信息；通过聚类算法，根据用户信息，对交易用户进行聚类，得到用户聚类结果；根据用户聚类结果、产品信息和已购产品的交易用户的用户信息，生成用户类型对应的预推荐产品信息；通过分类算法，根据用户聚类结果和用户类型对应的预推荐产品信息，对预先生成的目标用户的用户画像进行分类预测，生成目标推荐产品信息，对已购产品的交易数据对用户进行聚类分析，结合问卷应答以无侵入的方式嵌入原有的产品申购页面，通过完善用户特征、即时更新用户信息的方式精确将用户进行归类，结合用户偏好和产品收益，更加精准的预测用户的购买产品，更加快捷的促成交易，提高产品推荐精度，从而提升用户体验，进而提升产品的销售效率。In the scheme of the embodiment of the present invention, the transaction data of purchased products are obtained, and the transaction data includes product information and user information of transaction users of purchased products; through a clustering algorithm, transaction users are clustered according to user information to obtain user clustering results; based on the user clustering results, product information and user information of transaction users of purchased products, pre-recommended product information corresponding to the user type is generated; through a classification algorithm, based on the user clustering results and the pre-recommended product information corresponding to the user type, the user portrait of the pre-generated target user is classified and predicted to generate target recommended product information, the transaction data of purchased products is clustered and analyzed on the users, and the data is embedded in the original product purchase page in a non-invasive manner in combination with questionnaire answers, and users are accurately classified by improving user characteristics and updating user information in real time, and the user's purchase products are predicted more accurately in combination with user preferences and product benefits, so as to facilitate transactions more quickly and improve the accuracy of product recommendations, thereby improving user experience and further improving product sales efficiency.
上述实施例阐明的系统、装置、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机设备，具体的，计算机设备例如可以为个人计算机、膝上型计算机、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任何设备的组合。The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer device, and specifically, the computer device may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
本发明实施例提供了一种计算机设备，包括存储器和处理器，存储器用于存储包括程序指令的信息，处理器用于控制程序指令的执行，程序指令被处理器加载并执行时实现上述产品推荐方法的实施例的各步骤，具体描述可参见上述产品推荐方法的实施例。An embodiment of the present invention provides a computer device, including a memory and a processor, the memory is used to store information including program instructions, the processor is used to control the execution of the program instructions, and the program instructions, when loaded and executed by the processor, implement the steps of the embodiment of the above-mentioned product recommendation method. For a specific description, please refer to the embodiment of the above-mentioned product recommendation method.
下面参考图4，其示出了适于用来实现本申请实施例的计算机设备600的结构示意图。Reference is now made to FIG4 , which shows a schematic diagram of the structure of a computer device 600 suitable for implementing an embodiment of the present application.
如图4所示，计算机设备600包括中央处理单元(CPU)601，其可以根据存储在只读存储器(ROM)602中的程序或者从存储部分608加载到随机访问存储器(RAM)603中的程序而执行各种适当的工作和处理。在RAM603中，还存储有计算机设备600操作所需的各种程序和数据。CPU601、ROM602、以及RAM603通过总线604彼此相连。输入/输出(I/O)接口605也连接至总线604。As shown in FIG4 , the computer device 600 includes a central processing unit (CPU) 601, which can perform various appropriate operations and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage portion 608 into a random access memory (RAM) 603. Various programs and data required for the operation of the computer device 600 are also stored in the RAM 603. The CPU 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to the bus 604.
以下部件连接至I/O接口605：包括键盘、鼠标等的输入部分606；包括诸如阴极射线管(CRT)、液晶反馈器(LCD)等以及扬声器等的输出部分607；包括硬盘等的存储部分608；以及包括诸如LAN卡，调制解调器等的网络接口卡的通信部分609。通信部分609经由诸如因特网的网络执行通信处理。驱动器610也根据需要连接至I/O接口605。可拆卸介质611，诸如磁盘、光盘、磁光盘、半导体存储器等等，根据需要安装在驱动器610上，以便于从其上读出的计算机程序根据需要被安装如存储部分608。The following components are connected to the I/O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal feedback device (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I/O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed as needed as the storage section 608.
特别地，根据本发明的实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本发明的实施例包括一种计算机程序产品，其包括有形地包含在机器可读介质上的计算机程序，所述计算机程序包括用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信部分609从网络上被下载和安装，和/或从可拆卸介质611被安装。In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 609, and/or installed from the removable medium 611.
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
为了描述的方便，描述以上装置时以功能分为各种单元分别描述。当然，在实施本申请时可以把各单元的功能在同一个或多个软件和/或硬件中实现。For the convenience of description, the above device is described in terms of functions and is described separately in various units. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and/or hardware.
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The present invention is described with reference to the flowchart and/or block diagram of the method, device (system), and computer program product according to the embodiment of the present invention. It should be understood that each process and/or box in the flowchart and/or block diagram, as well as the combination of the process and/or box in the flowchart and/or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
本申请技术方案中对数据的获取、存储、使用、处理等均符合国家法律法规的相关规定。The acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.
需要说明的是，在本申请实施例中，可能提及某些软件、组件、模型等业界已有方案，应当将它们认为是示范性的，其目的仅仅是为了说明本申请技术方案实施中的可行性，但并不意味着申请人已经或者必然用到了该方案。It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.
本领域技术人员应明白，本申请的实施例可提供为方法、系统或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
本申请可以在由计算机执行的计算机可执行指令的一般上下文中描述，例如程序模块。一般地，程序模块包括执行特定任务或实现特定抽象数据类型的例程、程序、对象、组件、数据结构等等。也可以在分布式计算环境中实践本申请，在这些分布式计算环境中，由通过通信网络而被连接的远程处理设备来执行任务。在分布式计算环境中，程序模块可以位于包括存储设备在内的本地和远程计算机存储介质中。The present application may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于系统实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
以上所述仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410701484.6ACN118537136A (en) | 2024-05-31 | 2024-05-31 | Product recommendation method and device |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410701484.6ACN118537136A (en) | 2024-05-31 | 2024-05-31 | Product recommendation method and device |
| Publication Number | Publication Date |
|---|---|
| CN118537136Atrue CN118537136A (en) | 2024-08-23 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410701484.6APendingCN118537136A (en) | 2024-05-31 | 2024-05-31 | Product recommendation method and device |
| Country | Link |
|---|---|
| CN (1) | CN118537136A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119624584A (en)* | 2024-12-02 | 2025-03-14 | 中国平安人寿保险股份有限公司 | Product recommendation method and device, electronic device and storage medium |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119624584A (en)* | 2024-12-02 | 2025-03-14 | 中国平安人寿保险股份有限公司 | Product recommendation method and device, electronic device and storage medium |
| Publication | Publication Date | Title |
|---|---|---|
| CN108564286B (en) | Artificial intelligent financial wind-control credit assessment method and system based on big data credit investigation | |
| CN109063945B (en) | Value evaluation system-based 360-degree customer portrait construction method for electricity selling company | |
| CN112561598A (en) | Customer loss prediction and retrieval method and system based on customer portrait | |
| CN110956273A (en) | Credit scoring method and system integrating multiple machine learning models | |
| Chen | Classifying credit ratings for Asian banks using integrating feature selection and the CPDA-based rough sets approach | |
| CN108711107A (en) | Intelligent financing services recommend method and its system | |
| CN106156092A (en) | Data processing method and device | |
| Wang et al. | Advertisement click-through rate prediction using multiple criteria linear programming regression model | |
| ABIDAR et al. | Predicting customer segment changes to enhance customer retention: a case study for online retail using machine learning | |
| CN118780837A (en) | A method and system for constructing a personalized traffic user portrait application based on big data processing and multi-dimensional features | |
| CN117455529A (en) | Method and system for constructing user electricity consumption characteristic portrait based on big data technology | |
| Hu | Predicting and improving invoice-to-cash collection through machine learning | |
| CN118537136A (en) | Product recommendation method and device | |
| Liu et al. | Methods for aggregating investor sentiment from social media | |
| CN114626940A (en) | Data analysis method and device and electronic equipment | |
| CN113269610A (en) | Bank product recommendation method and device and storage medium | |
| CN118626910A (en) | Method, device and server for determining customer profile | |
| CN118504752A (en) | Determination method of transaction risk prediction model, transaction risk prediction method, device, equipment, storage medium and program product | |
| Rudnichenko et al. | Intelligent Data Clustering System for Searching Hidden Regularities in Financial Transactions. | |
| CN115205025A (en) | Risk account identification method and device, computer equipment and storage medium | |
| CN113988969A (en) | Collaborative filtering recommendation method based on RFM model | |
| Li et al. | Research on listed companies’ credit ratings, considering classification performance and interpretability | |
| CN113159966A (en) | Capital planning recommendation method, device, equipment and storage medium based on big data | |
| Sun et al. | Enhancing Credit Risk Prediction through an Ensemble of Explainable Model | |
| Bochkaryov et al. | Application of the ensemble clustering algorithm in solving the problem of segmentation of users taking into account their loyalty |
| 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 |