CN118606559A - Product recommendation method, device, equipment and storage medium - Google Patents

Abstract

Translated fromChinese

本申请涉及数据处理技术领域，尤其涉及一种产品推荐方法、装置、设备及存储介质。该方法通过对预设产品数据与客户数据进行分类，得到多类型数据；对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。该方法通过数据分类和标准化处理，生成初始客户画像，并更新客户画像以实时反映客户需求，最终基于目标客户画像推荐产品，提高了数据处理一致性和推荐精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果。

The present application relates to the field of data processing technology, and in particular to a product recommendation method, device, equipment and storage medium. The method obtains multiple types of data by classifying preset product data and customer data; standardizes the multiple types of data to obtain feature standard data; then generates an initial customer portrait based on the feature standard data and preset historical feature data; then obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; finally, determines the target recommended product based on the target customer portrait. The method generates an initial customer portrait through data classification and standardization, and updates the customer portrait to reflect customer needs in real time, and finally recommends products based on the target customer portrait, thereby improving data processing consistency and recommendation accuracy, ensuring that the recommended products are highly matched with customer needs, and improving the effect of product recommendations.

Description

Translated fromChinese

产品推荐方法、装置、设备及存储介质Product recommendation method, device, equipment and storage medium

技术领域Technical Field

本申请涉及数据处理技术领域，尤其涉及一种产品推荐方法、装置、设备及存储介质。The present application relates to the field of data processing technology, and in particular to a product recommendation method, device, equipment and storage medium.

背景技术Background Art

产品推荐过程应该围绕消费者行为展开，因此有必要想办法理解消费者，以方便交易的执行。目前，产品推荐的分析主要依靠将相关数据输入至Excel内，业务人员通过查看Excel内的营销数据，人工分析评价产品推荐业务效能，但是这种方式消耗了大量人力物力，对企业来说增加了大量的人工成本，并且人工不能根据Excel内的数据预测出合适的产品推荐策略，导致产品推荐的效果较差。The product recommendation process should revolve around consumer behavior, so it is necessary to find ways to understand consumers to facilitate the execution of transactions. At present, the analysis of product recommendations mainly relies on inputting relevant data into Excel. Business personnel manually analyze and evaluate the effectiveness of product recommendation business by viewing the marketing data in Excel. However, this method consumes a lot of manpower and material resources, increases a lot of labor costs for enterprises, and humans cannot predict appropriate product recommendation strategies based on the data in Excel, resulting in poor product recommendation results.

因此，如何提高产品推荐的准确度，提升产品推荐效果是目前亟需解决的一个问题。Therefore, how to improve the accuracy of product recommendations and enhance the effectiveness of product recommendations is a problem that needs to be solved urgently.

上述内容仅用于辅助理解本申请的技术方案，并不代表承认上述内容是现有技术。The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art.

发明内容Summary of the invention

本申请的主要目的在于提供了一种产品推荐方法、装置、设备及存储介质，旨在解决如何提高产品推荐的准确度，提升产品推荐效果的技术问题。The main purpose of this application is to provide a product recommendation method, device, equipment and storage medium, aiming to solve the technical problem of how to improve the accuracy of product recommendations and enhance the product recommendation effect.

为实现上述目的，本申请提供了一种产品推荐方法，所述方法包括以下步骤：To achieve the above object, the present application provides a product recommendation method, which comprises the following steps:

对预设产品数据与客户数据进行分类，得到多类型数据；Classify the preset product data and customer data to obtain multiple types of data;

对所述多类型数据进行标准化处理，得到特征标准数据；Performing standardization processing on the multi-type data to obtain characteristic standard data;

基于所述特征标准数据与预设历史特征数据，生成初始客户画像；Generate an initial customer profile based on the characteristic standard data and the preset historical characteristic data;

获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据更新所述初始客户画像，得到目标客户画像；Acquire customer interaction data, and update the initial customer profile based on the customer interaction data and customer potential demand data to obtain a target customer profile;

根据所述目标客户画像，确定目标推荐产品。Determine target recommended products based on the target customer portrait.

在一实施例中，所述对所述多类型数据进行标准化处理，得到特征标准数据的步骤，包括：In one embodiment, the step of performing standardization processing on the multi-type data to obtain characteristic standard data includes:

对所述多类型数据进行数据清洗，得到清洗数据，其中，所述数据清洗包括缺失值处理、重复值处理以及异常值处理中的一种或多种；Performing data cleaning on the multi-type data to obtain cleaned data, wherein the data cleaning includes one or more of missing value processing, duplicate value processing, and outlier processing;

对所述清洗数据进行归一化处理，得到归一化数据；Normalizing the cleaned data to obtain normalized data;

将所述归一化数据整合至预设格式化表中，得到所述特征标准数据。The normalized data is integrated into a preset formatting table to obtain the characteristic standard data.

在一实施例中，所述基于所述特征标准数据与预设历史特征数据，生成初始客户画像的步骤，包括：In one embodiment, the step of generating an initial customer profile based on the characteristic standard data and the preset historical characteristic data includes:

提取所述预设历史特征数据中与所述特征标准数据的相似度超过预设相似度阈值的数据，得到待融合数据；Extracting data from the preset historical feature data whose similarity with the feature standard data exceeds a preset similarity threshold to obtain data to be fused;

将所述待融合数据与所述特征标准数据进行加权融合，得到融合数据；Performing weighted fusion on the data to be fused and the characteristic standard data to obtain fused data;

对所述融合数据进行特征提取，并基于特征提取结果与预设画像生成模型，生成所述初始客户画像。Feature extraction is performed on the fused data, and the initial customer portrait is generated based on the feature extraction result and a preset portrait generation model.

在一实施例中，所述获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据更新所述初始客户画像，得到目标客户画像的步骤，包括：In one embodiment, the step of acquiring customer interaction data, and updating the initial customer profile based on the customer interaction data and customer potential demand data to obtain a target customer profile includes:

对所述初始客户画像进行识别，确定客户潜在需求数据；Identify the initial customer portrait and determine the customer's potential demand data;

获取所述客户互动数据，并对所述客户互动数据与所述客户潜在需求数据进行所述标准化处理，得到补充特征数据；Acquiring the customer interaction data, and performing the standardization process on the customer interaction data and the customer potential demand data to obtain supplementary feature data;

基于所述补充特征数据，调整所述预设画像生成模型的参数与权重；Based on the supplementary feature data, adjusting the parameters and weights of the preset portrait generation model;

根据调整后的所述预设画像生成模型、所述补充特征数据以及所述特征提取结果，得到所述目标客户画像。The target customer portrait is obtained based on the adjusted preset portrait generation model, the supplementary feature data and the feature extraction result.

在一实施例中，所述对所述初始客户画像进行识别，确定客户潜在需求数据的步骤，包括：In one embodiment, the step of identifying the initial customer profile and determining the customer's potential demand data includes:

基于预设特征识别算法，对所述初始客户画像进行特征识别，得到行为特征、关系特征以及偏好特征；Based on a preset feature recognition algorithm, feature recognition is performed on the initial customer portrait to obtain behavioral features, relationship features, and preference features;

根据所述行为特征、所述关系特征以及所述偏好特征，对客户行为与客户需求进行预测，得到行为预测数据与需求预测数据；Predicting customer behavior and customer demand based on the behavior characteristics, the relationship characteristics, and the preference characteristics to obtain behavior prediction data and demand prediction data;

对所述行为预测数据与所述需求预测数据进行需求概率评估，并将超过预设需求概率阈值的概率评估结果对应的所述行为预测数据或所述需求预测数据作为所述客户潜在需求数据。A demand probability evaluation is performed on the behavior prediction data and the demand prediction data, and the behavior prediction data or the demand prediction data corresponding to the probability evaluation result exceeding a preset demand probability threshold is used as the customer potential demand data.

在一实施例中，所述根据所述目标客户画像，确定目标推荐产品的步骤，包括：In one embodiment, the step of determining a target recommended product according to the target customer portrait includes:

对所述目标客户画像进行关键特征提取，得到关键特征数据；Extract key features of the target customer portrait to obtain key feature data;

将所述关键特征数据与所述预设产品数据进行归一化处理，得到归一化关键特征数据与归一化产品数据；Normalizing the key feature data and the preset product data to obtain normalized key feature data and normalized product data;

将所述归一化关键特征数据与所述归一化产品数据中的各产品特征数据进行匹配；Matching the normalized key feature data with each product feature data in the normalized product data;

根据匹配度高于预设匹配度阈值对应的产品特征数据，确定所述目标推荐产品。The target recommended product is determined according to the product feature data corresponding to the matching degree being higher than the preset matching degree threshold.

在一实施例中，在所述根据所述目标客户画像，确定目标推荐产品的步骤之后，还包括：In one embodiment, after the step of determining the target recommended product according to the target customer portrait, the following step is further included:

获取购买详情信息，并判断所述购买详情信息中是否包括所述目标推荐产品；Acquire purchase detail information, and determine whether the purchase detail information includes the target recommended product;

若不包括，获取客户反馈信息，并根据所述客户反馈信息，确定购买决策影响因素；If not, obtain customer feedback information and determine the factors influencing the purchase decision based on the customer feedback information;

基于所述购买决策影响因素与所述关键特征数据，生成产品推荐优化策略。Based on the purchase decision influencing factors and the key feature data, a product recommendation optimization strategy is generated.

此外，为实现上述目的，本申请还提出一种产品推荐装置，所述产品推荐装置包括：In addition, to achieve the above purpose, the present application also proposes a product recommendation device, which includes:

分类模块，用于对预设产品数据与客户数据进行分类，得到多类型数据；A classification module is used to classify preset product data and customer data to obtain multiple types of data;

标准化模块，用于对所述多类型数据进行标准化处理，得到特征标准数据；A standardization module, used for performing standardization processing on the multi-type data to obtain characteristic standard data;

初始画像生成模块，用于基于所述特征标准数据与预设历史特征数据，生成初始客户画像；An initial profile generation module, used to generate an initial customer profile based on the characteristic standard data and preset historical characteristic data;

画像更新模块，用于获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据，更新所述初始客户画像，得到目标客户画像；A portrait updating module is used to obtain customer interaction data, and based on the customer interaction data and customer potential demand data, update the initial customer portrait to obtain a target customer portrait;

产品推荐模块，用于根据所述目标客户画像，确定目标推荐产品。The product recommendation module is used to determine target recommended products based on the target customer portrait.

此外，为实现上述目的，本申请还提出一种产品推荐设备，所述设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的产品推荐程序，所述产品推荐程序配置为实现如上文所述的产品推荐方法的步骤。In addition, to achieve the above-mentioned purpose, the present application also proposes a product recommendation device, which includes: a memory, a processor, and a product recommendation program stored in the memory and executable on the processor, and the product recommendation program is configured to implement the steps of the product recommendation method described above.

此外，为实现上述目的，本申请还提出一种存储介质，所述存储介质上存储有产品推荐程序，所述产品推荐程序被处理器执行时实现如上文所述的产品推荐的步骤。In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, on which a product recommendation program is stored, and when the product recommendation program is executed by a processor, the steps of product recommendation as described above are implemented.

本申请通过先对预设产品数据与客户数据进行分类，得到多类型数据；接着对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。本申请通过数据分类和标准化处理，生成初始客户画像，并动态更新客户画像以实时反映客户需求，最终基于目标客户画像精准推荐产品，提高了数据处理一致性和推荐精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果。This application first classifies the preset product data and customer data to obtain multiple types of data; then standardizes the multiple types of data to obtain feature standard data; then generates an initial customer portrait based on the feature standard data and preset historical feature data; then obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain the target customer portrait; finally, determines the target recommended product based on the target customer portrait. This application generates an initial customer portrait through data classification and standardization, and dynamically updates the customer portrait to reflect customer needs in real time, and finally accurately recommends products based on the target customer portrait, which improves data processing consistency and recommendation accuracy, ensures that the recommended products are highly matched with customer needs, and improves the effect of product recommendations.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

图1为本申请产品推荐方法第一实施例的流程示意图；FIG1 is a flow chart of a first embodiment of a product recommendation method of the present application;

图2为本申请产品推荐方法第二实施例中一子流程示意图；FIG2 is a schematic diagram of a sub-process in the second embodiment of the product recommendation method of the present application;

图3为本申请产品推荐方法第二实施例中又一子流程示意图；FIG3 is a schematic diagram of another sub-process in the second embodiment of the product recommendation method of the present application;

图4为本申请产品推荐方法第三实施例中一子流程示意图；FIG4 is a schematic diagram of a sub-process in the third embodiment of the product recommendation method of the present application;

图5为本申请产品推荐方法第三实施例中又一子流程示意图；FIG5 is a schematic diagram of another sub-process in the third embodiment of the product recommendation method of the present application;

图6为本申请产品推荐方法一实施例中产品推荐方法系统架构示例图；FIG6 is a diagram showing an example of a system architecture of a product recommendation method in accordance with an embodiment of a product recommendation method of the present application;

图7为本申请产品推荐方法一实施例中数据处理示例图；FIG. 7 is a diagram showing an example of data processing in an embodiment of a product recommendation method of the present application;

图8为本申请实施例产品推荐装置的模块结构示意图；FIG8 is a schematic diagram of the module structure of the product recommendation device according to an embodiment of the present application;

图9为本申请实施例中产品推荐方法涉及的硬件运行环境的设备结构示意图。FIG9 is a schematic diagram of the device structure of the hardware operating environment involved in the product recommendation method in an embodiment of the present application.

本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings.

具体实施方式DETAILED DESCRIPTION

应当理解，此处所描述的具体实施例仅用以解释本申请，并不用于限定本申请。It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

为了更好的理解本申请的技术方案，下面将结合说明书附图以及具体的实施方式进行详细的说明。In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

需要说明的是，产品推荐过程应该围绕消费者行为展开，因此有必要想办法理解消费者，以方便交易的执行。目前，产品推荐的分析主要依靠将相关数据输入至Excel内，业务人员通过查看Excel内的营销数据，人工分析评价产品推荐业务效能，但是这种方式消耗了大量人力物力，对企业来说增加了大量的人工成本，并且人工不能根据Excel内的数据预测出合适的产品推荐策略，导致产品推荐的效果较差。因此，如何提高产品推荐的准确度，提升产品推荐效果是目前亟需解决的一个问题。It should be noted that the product recommendation process should revolve around consumer behavior, so it is necessary to find ways to understand consumers to facilitate the execution of transactions. At present, the analysis of product recommendations mainly relies on inputting relevant data into Excel. Business personnel manually analyze and evaluate the business effectiveness of product recommendations by viewing the marketing data in Excel. However, this method consumes a lot of manpower and material resources, increases a lot of labor costs for enterprises, and humans cannot predict appropriate product recommendation strategies based on the data in Excel, resulting in poor product recommendation results. Therefore, how to improve the accuracy of product recommendations and improve the effectiveness of product recommendations is a problem that needs to be solved urgently.

本申请的主要解决方案是：通过先对预设产品数据与客户数据进行分类，得到多类型数据；接着对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。The main solution of this application is: by first classifying the preset product data and customer data to obtain multi-type data; then standardizing the multi-type data to obtain feature standard data; then generating an initial customer portrait based on the feature standard data and preset historical feature data; then obtaining customer interaction data, and updating the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; finally, determining the target recommended product based on the target customer portrait.

本申请通过数据分类和标准化处理，生成初始客户画像，并动态更新客户画像以实时反映客户需求，最终基于目标客户画像精准推荐产品，提高了数据处理一致性和产品推荐的精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果，提升了营销效能。This application generates an initial customer portrait through data classification and standardization, and dynamically updates the customer portrait to reflect customer needs in real time. It ultimately recommends products accurately based on the target customer portrait, improves data processing consistency and the accuracy of product recommendations, ensures that recommended products are highly matched with customer needs, improves the effectiveness of product recommendations, and improves marketing efficiency.

需要说明的是，本实施例方法的执行主体可以是具有数据处理、网络通信以及程序运行功能的计算服务设备，也可以是具有相同或相似功能的上述产品推荐设备。本实施例及下述各实施例将以产品推荐设备为例进行说明。It should be noted that the execution subject of the method of this embodiment can be a computing service device with data processing, network communication and program running functions, or it can be the above-mentioned product recommendation device with the same or similar functions. This embodiment and the following embodiments will be described by taking the product recommendation device as an example.

基于此，提出本申请产品推荐方法的第一实施例，请参照图1，图1为本申请产品推荐方法第一实施例的流程示意图。Based on this, a first embodiment of the product recommendation method of the present application is proposed. Please refer to FIG. 1 , which is a flow chart of the first embodiment of the product recommendation method of the present application.

在本实施例中，该产品推荐方法包括以下步骤：In this embodiment, the product recommendation method includes the following steps:

S1：对预设产品数据与客户数据进行分类，得到多类型数据；S1: Classify the preset product data and customer data to obtain multiple types of data;

具体的，收集预设产品数据指的是从银行内部系统或其他数据源收集关于银行产品的信息。这些数据包括产品名称、类别、特征、价格、适用客户群、历史销售记录等。收集客户数据指的是收集企业所属行业信息、企业所在地区信息、企业注册资本信息等企业的基本信息。Specifically, collecting preset product data refers to collecting information about bank products from the bank's internal system or other data sources. This data includes product name, category, features, price, applicable customer group, historical sales records, etc. Collecting customer data refers to collecting basic information about the company, such as the industry to which the company belongs, the region where the company is located, and the registered capital of the company.

需要说明的是，本实施例中进行产品推荐的客户指的是对公客户。It should be noted that, in this embodiment, the customers for whom product recommendations are made refer to corporate customers.

进一步的，多类型数据包括标签数据、关系数据以及非结构化数据。标签数据为结构化的数据，例如数值、类别等，便于直接分析和处理。包括企业的基本信息、账户余额、交易频率等。关系数据为描述企业与产品或其他企业之间的关系，例如企业购买过的产品、企业与客户经理的关系等。非结构化数据为文本、图像等非结构化数据，例如反馈意见、聊天记录等。在分类后，将分类后的多类型数据存储在统一的数据库中，确保数据的组织结构清晰，便于后续的处理和分析。Furthermore, multi-type data includes label data, relational data, and unstructured data. Label data is structured data, such as numerical values, categories, etc., which is easy to directly analyze and process. It includes basic information of the enterprise, account balance, transaction frequency, etc. Relational data describes the relationship between the enterprise and products or other enterprises, such as products purchased by the enterprise, the relationship between the enterprise and the account manager, etc. Unstructured data is unstructured data such as text and images, such as feedback, chat records, etc. After classification, the classified multi-type data is stored in a unified database to ensure that the organizational structure of the data is clear and convenient for subsequent processing and analysis.

分类后的数据结构清晰，便于标准化处理和后续分析，减少了处理异构数据的复杂性。不同类型的数据分类存储，能够更有效地利用各类数据，提供更丰富的对公客户信息和产品特征，为精准推荐奠定基础。通过将结构化、关系和非结构化数据进行分类存储，可以进行多维度的数据分析和挖掘，提升画像的准确性和推荐系统的精准度。The structure of classified data is clear, which is convenient for standardized processing and subsequent analysis, and reduces the complexity of processing heterogeneous data. Different types of data classification and storage can make more effective use of various types of data, provide richer corporate customer information and product features, and lay the foundation for accurate recommendations. By classifying and storing structured, relational, and unstructured data, multi-dimensional data analysis and mining can be performed to improve the accuracy of the portrait and the accuracy of the recommendation system.

S2：对所述多类型数据进行标准化处理，得到特征标准数据；S2: performing standardization processing on the multi-type data to obtain characteristic standard data;

具体的，标准化处理包括数据清洗、数据格式统一、数据归一化、特征提取与生成以及特征合并与存储。数据清洗指的是处理数据中的缺失值、重复值和异常值，确保数据质量的过程。数据归一化指的是将数据转换到统一尺度的过程，以便于后续分析和建模。特征提取指的是从原始数据中提取有用特征的过程。特征生成指的是根据业务需求生成新的特征的过程。Specifically, standardization processing includes data cleaning, data format unification, data normalization, feature extraction and generation, and feature merging and storage. Data cleaning refers to the process of processing missing values, duplicate values, and outliers in data to ensure data quality. Data normalization refers to the process of converting data to a uniform scale for subsequent analysis and modeling. Feature extraction refers to the process of extracting useful features from raw data. Feature generation refers to the process of generating new features based on business needs.

通过数据清洗处理缺失值、重复值和异常值，确保数据的完整性和准确性，提高数据质量。统一数据格式和类型，确保不同来源和不同类型的数据可以无缝整合，提高数据的一致性。归一化和编码处理后的数据便于进行后续的机器学习和数据挖掘，提高模型的稳定性和准确性。提取和生成有用的特征，丰富了数据的特征信息，为后续的客户画像生成和精准推荐提供了坚实的基础。By cleaning the data, missing values, duplicate values and outliers are processed to ensure the integrity and accuracy of the data and improve the data quality. The data format and type are unified to ensure that data from different sources and different types can be seamlessly integrated and improve the consistency of the data. The normalized and encoded data facilitates subsequent machine learning and data mining, improving the stability and accuracy of the model. The extraction and generation of useful features enriches the feature information of the data and provides a solid foundation for the subsequent generation of customer portraits and accurate recommendations.

S3：基于所述特征标准数据与预设历史特征数据，生成初始客户画像；S3: Generate an initial customer profile based on the characteristic standard data and the preset historical characteristic data;

具体的，从数据库中提取经过标准化处理的客户特征数据，这些数据包括企业的基本信息、行为特征、关系特征等。从历史数据库中提取预设的客户历史特征数据，这些数据通常包括企业的历史交易记录、过去的购买行为、意见反馈等。确保提取的特征标准数据与预设历史特征数据中的客户标识(如客户ID)能够正确匹配。将特征标准数据与预设历史特征数据进行融合，形成一个包含客户当前特征和历史特征的综合数据集。Specifically, standardized customer feature data is extracted from the database. These data include basic information, behavioral features, relationship features, etc. of the enterprise. Preset customer historical feature data is extracted from the historical database. These data usually include the enterprise's historical transaction records, past purchase behaviors, feedback, etc. Ensure that the extracted feature standard data and the customer identifier (such as customer ID) in the preset historical feature data can be correctly matched. The feature standard data is merged with the preset historical feature data to form a comprehensive data set containing the customer's current features and historical features.

进一步的，从融合后的数据集中选择对客户画像生成有重要影响的关键特征，如财务信息、运营信息、企业客户群体信息、法律合规信息等。使用数据分析和统计方法(如相关分析、主成分分析等)分析选定特征的分布和关系，确保数据的准确性和有效性。应用聚类算法(如K-means、层次聚类等)将客户分成不同的群体，每个群体代表一类具有相似特征和行为模式的客户。通过分析每个客户群体的特征，验证分群结果的合理性，确保客户分群能够有效反映客户的差异性。基于特征提取和分群结果，构建客户画像模型。模型可以采用规则引擎或机器学习算法(如决策树、随机森林等)来生成客户画像。根据画像模型生成初始客户画像，包括客户的基本信息、行为特征、关系特征、偏好特征等。Furthermore, key features that have an important impact on the generation of customer portraits are selected from the fused data set, such as financial information, operational information, corporate customer group information, legal compliance information, etc. Use data analysis and statistical methods (such as correlation analysis, principal component analysis, etc.) to analyze the distribution and relationship of the selected features to ensure the accuracy and validity of the data. Apply clustering algorithms (such as K-means, hierarchical clustering, etc.) to divide customers into different groups, each of which represents a class of customers with similar characteristics and behavior patterns. By analyzing the characteristics of each customer group, verify the rationality of the grouping results to ensure that customer grouping can effectively reflect the differences of customers. Based on feature extraction and grouping results, build a customer portrait model. The model can use a rule engine or machine learning algorithm (such as decision tree, random forest, etc.) to generate customer portraits. Generate an initial customer portrait based on the portrait model, including the customer's basic information, behavioral characteristics, relationship characteristics, preference characteristics, etc.

可选的，将生成的初始客户画像存储在数据库中，便于后续查询和应用。初始客户画像可以应用于客户需求预测、精准营销、个性化推荐等多个场景，提高业务决策的精准度和效率。Optionally, the generated initial customer portrait is stored in the database for subsequent query and application. The initial customer portrait can be applied to multiple scenarios such as customer demand forecasting, precision marketing, and personalized recommendations to improve the accuracy and efficiency of business decisions.

需要说明的是，特征标准数据为经过标准化处理后的客户特征数据，具有一致性和可用性。预设历史特征数据为历史数据库中预设的客户特征数据，通常包括客户的历史交易记录和行为数据。客户画像是基于客户特征数据生成的客户描述，反映客户的基本信息、行为特征和偏好特征。It should be noted that the characteristic standard data is the customer characteristic data that has been standardized and has consistency and availability. The preset historical characteristic data is the customer characteristic data preset in the historical database, which usually includes the customer's historical transaction records and behavior data. The customer portrait is a customer description generated based on the customer characteristic data, reflecting the customer's basic information, behavior characteristics and preference characteristics.

结合当前特征数据和历史特征数据，生成的客户画像更加全面和准确，能够反映客户的全貌和行为变化。通过聚类分析，将客户分成不同的群体，有助于识别客户的差异性和个性化需求，提高营销和服务的针对性。采用数据分析和机器学习算法，快速生成初始客户画像，提升画像生成的效率和准确性。生成的初始客户画像可以应用于精准营销和个性化推荐，提高客户满意度和营销成功率。Combining current feature data with historical feature data, the generated customer portrait is more comprehensive and accurate, and can reflect the overall picture of the customer and changes in behavior. Through cluster analysis, customers can be divided into different groups, which helps to identify customer differences and personalized needs, and improve the targeting of marketing and services. Using data analysis and machine learning algorithms, the initial customer portrait is quickly generated, improving the efficiency and accuracy of portrait generation. The generated initial customer portrait can be applied to precision marketing and personalized recommendations to improve customer satisfaction and marketing success rate.

S4：获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据更新所述初始客户画像，得到目标客户画像。S4: Acquire customer interaction data, and update the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait.

需要说明的是，客户互动数据包括在线互动数据、线下互动数据等。在线互动数据指的是客户在银行网站、移动应用、电子邮件等在线渠道的互动数据，包括浏览记录、点击记录、搜索查询、在线咨询等。线下互动数据指的是客户在银行网点或通过电话客服的互动数据，包括拜访记录、电话记录、面谈记录。It should be noted that customer interaction data includes online interaction data, offline interaction data, etc. Online interaction data refers to customer interaction data on online channels such as bank websites, mobile applications, and emails, including browsing records, click records, search queries, online consultations, etc. Offline interaction data refers to customer interaction data at bank branches or through telephone customer service, including visit records, phone records, and interview records.

需要特别说明的是，本申请涉及到的客户互动数据等相关数据(例如，在线互动数据、线下互动数据等)，均是在获得用户的许可或者同意后获取到的；也就是说，当本申请运用到具体产品或技术中时，需要获得用户许可来实现相关数据的获取和处理，且相关数据的处理需要遵守相关国家和地区的相关法律法规和监管标准。It should be noted that the customer interaction data and other relevant data involved in this application (for example, online interaction data, offline interaction data, etc.) are all obtained after obtaining the user's permission or consent; that is, when this application is applied to specific products or technologies, it is necessary to obtain user permission to obtain and process the relevant data, and the processing of the relevant data needs to comply with the relevant laws, regulations and regulatory standards of the relevant countries and regions.

具体的，处理互动数据中的缺失值、重复值和异常值，确保数据质量。统一数据格式，确保不同来源的互动数据可以整合在一起进行分析。将客户的在线互动数据和线下互动数据整合到一个统一的数据库中。将整合的互动数据与初始客户画像进行关联，确保所有互动数据都与正确的客户匹配。基于初始客户画像和互动数据，识别客户的潜在需求。使用机器学习模型(如协同过滤、决策树等)预测客户的潜在需求。将识别出的客户潜在需求进行分类，形成客户需求特征数据。Specifically, handle missing values, duplicate values, and outliers in the interaction data to ensure data quality. Unify data formats to ensure that interaction data from different sources can be integrated for analysis. Integrate customers' online and offline interaction data into a unified database. Associate the integrated interaction data with the initial customer profile to ensure that all interaction data matches the correct customer. Identify customers' potential needs based on the initial customer profile and interaction data. Use machine learning models (such as collaborative filtering, decision trees, etc.) to predict customers' potential needs. Classify the identified potential needs of customers to form customer demand feature data.

进一步的，将新的互动数据和潜在需求数据输入客户画像模型，更新初始客户画像。从新的互动数据中提取关键特征，补充到客户画像中。根据新的特征数据，优化客户画像模型，确保画像的准确性和实时性。基于更新后的客户画像模型，生成目标客户画像。目标客户画像应包含最新的客户特征、行为模式和需求特征。通过客户经理反馈、客户行为观察等方式，验证目标客户画像的准确性，确保其能够准确反映客户的最新需求和行为。Furthermore, the new interaction data and potential demand data are input into the customer portrait model to update the initial customer portrait. Key features are extracted from the new interaction data and added to the customer portrait. Based on the new feature data, the customer portrait model is optimized to ensure the accuracy and real-time nature of the portrait. Based on the updated customer portrait model, the target customer portrait is generated. The target customer portrait should contain the latest customer characteristics, behavior patterns, and demand characteristics. The accuracy of the target customer portrait is verified through feedback from account managers, customer behavior observation, etc., to ensure that it can accurately reflect the customer's latest needs and behaviors.

需要说明的是，客户互动数据指的是客户在各种渠道与银行进行互动时产生的数据。客户潜在需求数据是基于客户历史行为和特征预测出的客户可能的需求数据。目标客户画像为在初始客户画像的基础上，结合最新的客户互动数据和潜在需求数据动态更新后生成的客户画像，能够准确反映客户的当前需求和行为模式。It should be noted that customer interaction data refers to the data generated when customers interact with banks through various channels. Customer potential demand data refers to the customer's possible demand data predicted based on the customer's historical behavior and characteristics. The target customer portrait is a customer portrait generated based on the initial customer portrait and dynamically updated with the latest customer interaction data and potential demand data, which can accurately reflect the customer's current needs and behavior patterns.

动态更新客户画像，能够实时反映客户的最新需求和行为变化，确保画像的时效性和准确性。结合最新的互动数据和潜在需求数据，优化客户画像模型，提升画像的准确性和全面性。生成的目标客户画像为精准营销和个性化服务提供了可靠的数据基础，提高客户满意度和营销成功率。实时更新和优化客户画像，使银行能够更快速地响应客户需求和市场变化，提升业务决策的敏捷性和效果。Dynamically updating customer portraits can reflect the latest customer needs and behavior changes in real time, ensuring the timeliness and accuracy of the portraits. Combining the latest interaction data and potential demand data, optimize the customer portrait model to improve the accuracy and comprehensiveness of the portrait. The generated target customer portrait provides a reliable data foundation for precision marketing and personalized services, improving customer satisfaction and marketing success rate. Real-time updating and optimization of customer portraits enables banks to respond to customer needs and market changes more quickly, and improve the agility and effectiveness of business decisions.

S5：根据所述目标客户画像，确定目标推荐产品。S5: Determine target recommended products based on the target customer portrait.

具体的，从目标客户画像中提取客户的关键特征，包括基本信息(如企业名称、注册资金、行业类别、注册地址)、行为特征(如交易频率、购买历史)、偏好特征(如产品喜好、兴趣爱好)等。根据提取的特征，将客户分为不同的群体(如高价值客户、潜在客户、频繁购买客户等)，以便为每个群体制定不同的推荐策略。收集银行所有产品的详细信息，包括产品名称、类别、功能、价格、适用客户群、历史销售记录等。从产品数据中提取关键特征，确保能够与客户的需求和偏好进行匹配。例如，提取产品的风险等级、收益率、适用场景等。Specifically, extract the key features of the customer from the target customer portrait, including basic information (such as company name, registered capital, industry category, registered address), behavioral characteristics (such as transaction frequency, purchase history), preference characteristics (such as product preferences, interests and hobbies), etc. According to the extracted features, divide the customers into different groups (such as high-value customers, potential customers, frequent purchasers, etc.) so as to formulate different recommendation strategies for each group. Collect detailed information on all bank products, including product name, category, function, price, applicable customer group, historical sales records, etc. Extract key features from product data to ensure that they can match customer needs and preferences. For example, extract the product's risk level, rate of return, applicable scenarios, etc.

进一步的，构建基于业务规则的推荐引擎，根据预设的规则(如客户偏好、历史购买行为、产品相似性等)匹配产品。使用机器学习算法(如协同过滤、矩阵分解、深度学习等)进行产品推荐。模型可以基于客户的历史行为、相似客户的购买记录、产品相似性等进行推荐。Furthermore, a recommendation engine based on business rules is built to match products according to preset rules (such as customer preferences, historical purchase behavior, product similarity, etc.). Use machine learning algorithms (such as collaborative filtering, matrix decomposition, deep learning, etc.) to make product recommendations. The model can make recommendations based on the customer's historical behavior, the purchase records of similar customers, product similarity, etc.

可选的，根据目标客户画像和产品匹配结果，生成推荐产品列表。推荐列表可以按相关性、受欢迎程度、客户偏好等进行排序。为每个推荐的产品生成推荐理由，解释为什么推荐该产品，增加客户对推荐结果的信任度和接受度。使用图表、图形等可视化工具展示推荐方案，使客户能够直观地看到推荐的产品及其特征。在银行的官网、移动应用等平台上设计友好的用户界面，展示推荐产品和理由，提升客户体验。Optionally, generate a list of recommended products based on the target customer profile and product matching results. The recommendation list can be sorted by relevance, popularity, customer preference, etc. Generate a recommendation reason for each recommended product, explaining why the product is recommended, and increasing customer trust and acceptance of the recommendation results. Use visualization tools such as charts and graphs to display the recommendation plan so that customers can intuitively see the recommended products and their features. Design a friendly user interface on the bank's official website, mobile application and other platforms to display recommended products and reasons and improve customer experience.

进一步的，收集客户对推荐产品的反馈，了解客户的满意度和实际需求。反馈方式可以包括问卷调查、在线评价、客户经理的反馈等。分析客户对推荐产品的行为数据，如点击、购买、浏览等，评估推荐效果。根据客户的反馈和行为数据，不断优化推荐算法和规则，提升推荐模型的准确性和效果。Furthermore, collect customer feedback on recommended products to understand customer satisfaction and actual needs. Feedback methods can include questionnaires, online reviews, feedback from account managers, etc. Analyze customer behavior data on recommended products, such as clicks, purchases, and browsing, to evaluate the effectiveness of recommendations. Based on customer feedback and behavior data, continuously optimize recommendation algorithms and rules to improve the accuracy and effectiveness of recommendation models.

基于客户的最新画像，智能匹配和推荐最适合客户需求的产品，确保推荐的产品与客户需求高度匹配，提高客户满意度和营销成功率。通过生成个性化的推荐列表和推荐理由，提升客户的个性化体验，增强客户对银行服务的信任和粘性。收集客户反馈和行为数据，不断优化推荐算法和规则，确保推荐系统能够实时适应客户需求的变化，提高推荐效果和精准度。自动化和智能化的推荐流程减少了客户经理的工作量，提升了业务决策的效率和效果，帮助银行更快速地响应市场和客户需求。Based on the latest customer profile, intelligently match and recommend products that best suit customer needs, ensure that recommended products are highly matched with customer needs, and improve customer satisfaction and marketing success rate. By generating personalized recommendation lists and recommendation reasons, customers' personalized experience is enhanced, and their trust and stickiness to banking services are strengthened. Customer feedback and behavior data are collected, and recommendation algorithms and rules are continuously optimized to ensure that the recommendation system can adapt to changes in customer needs in real time, and improve the effectiveness and accuracy of recommendations. The automated and intelligent recommendation process reduces the workload of account managers, improves the efficiency and effectiveness of business decision-making, and helps banks respond to market and customer needs more quickly.

本实施例通过先对预设产品数据与客户数据进行分类，得到多类型数据；接着对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。本实施例通过数据分类和标准化处理，生成初始客户画像，并动态更新客户画像以实时反映客户需求，最终基于目标客户画像精准推荐产品，提高了数据处理一致性和推荐精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果，提升了营销效能。This embodiment first classifies the preset product data and customer data to obtain multiple types of data; then standardizes the multiple types of data to obtain feature standard data; then generates an initial customer portrait based on the feature standard data and the preset historical feature data; then obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; finally, determines the target recommended product based on the target customer portrait. This embodiment generates an initial customer portrait through data classification and standardization, and dynamically updates the customer portrait to reflect customer needs in real time, and finally accurately recommends products based on the target customer portrait, thereby improving data processing consistency and recommendation accuracy, ensuring that the recommended products are highly matched with customer needs, improving the effect of product recommendations, and improving marketing efficiency.

基于上述第一实施例，提出本申请产品推荐方法的第二实施例。请参阅图2，图2为本申请产品推荐方法第二实施例中一子流程示意图。Based on the above first embodiment, a second embodiment of the product recommendation method of the present application is proposed. Please refer to Figure 2, which is a schematic diagram of a sub-flow in the second embodiment of the product recommendation method of the present application.

如图2所示，在本实施例中，步骤S2包括：As shown in FIG. 2 , in this embodiment, step S2 includes:

S21：对所述多类型数据进行数据清洗，得到清洗数据，其中，所述数据清洗包括缺失值处理、重复值处理以及异常值处理中的一种或多种；S21: performing data cleaning on the multi-type data to obtain cleaned data, wherein the data cleaning includes one or more of missing value processing, duplicate value processing, and outlier processing;

S22：对所述清洗数据进行归一化处理，得到归一化数据；S22: performing normalization processing on the cleaned data to obtain normalized data;

S23：将所述归一化数据整合至预设格式化表中，得到所述特征标准数据。S23: Integrate the normalized data into a preset formatting table to obtain the characteristic standard data.

具体的，扫描数据集，识别包含缺失值的记录。通过删除缺失值记录、填补缺失值或插值法来对缺失值进行处理。扫描数据集，识别重复记录，删除重复记录，确保每条记录的唯一性。使用统计方法(如箱线图)或机器学习方法(如孤立森林算法)识别异常值，删除识别出的异常值记录。Specifically, scan the data set to identify records containing missing values. Handle missing values by deleting missing value records, filling missing values, or interpolation. Scan the data set to identify duplicate records, delete duplicate records, and ensure the uniqueness of each record. Use statistical methods (such as box plots) or machine learning methods (such as the isolation forest algorithm) to identify outliers and delete the identified outlier records.

进一步的，将数值型数据缩放到统一的范围内，例如[0,1]范围，将类别型数据转换为数值型数据。将处理过的数值型数据和类别型数据整合到一个统一的数据表中。根据预设的格式化标准，确保数据表中所有字段的格式一致，包括数据类型、列名规范等。将整合后的特征标准数据存储到数据库或文件中，便于后续分析和使用。Furthermore, the numerical data is scaled to a uniform range, such as [0,1], and the categorical data is converted to numerical data. The processed numerical data and categorical data are integrated into a unified data table. According to the preset formatting standards, the format of all fields in the data table is ensured to be consistent, including data type, column name specifications, etc. The integrated feature standard data is stored in a database or file for subsequent analysis and use.

需要说明的是，数据清洗指的是处理数据中的缺失值、重复值和异常值，确保数据质量的过程。归一化处理指的是将数据缩放到统一尺度的过程，以便于后续分析和建模。预设格式化表为按照预设标准，统一格式的数据表。It should be noted that data cleaning refers to the process of dealing with missing values, duplicate values, and outliers in the data to ensure data quality. Normalization refers to the process of scaling data to a uniform scale to facilitate subsequent analysis and modeling. The preset formatted table is a data table in a uniform format according to preset standards.

数据清洗处理缺失值、重复值和异常值，确保数据的完整性和准确性，提高数据质量。归一化处理将数据缩放到统一尺度，统一了数据格式，确保数据的一致性和可用性。处理后的特征标准数据格式统一，便于进行后续的机器学习和数据挖掘，提高模型的稳定性和准确性。Data cleaning processes missing values, duplicate values, and outliers to ensure data integrity and accuracy and improve data quality. Normalization scales data to a uniform scale, unifies data formats, and ensures data consistency and availability. The processed feature standard data format is unified, which facilitates subsequent machine learning and data mining, and improves the stability and accuracy of the model.

基于上述第一实施例，在本实施例中，步骤S3包括：Based on the above first embodiment, in this embodiment, step S3 includes:

S31：提取所述预设历史特征数据中与所述特征标准数据的相似度超过预设相似度阈值的数据，得到待融合数据；S31: extracting data from the preset historical feature data whose similarity with the feature standard data exceeds a preset similarity threshold, to obtain data to be fused;

S32：将所述待融合数据与所述特征标准数据进行加权融合，得到融合数据；S32: performing weighted fusion on the data to be fused and the characteristic standard data to obtain fused data;

S33：对所述融合数据进行特征提取，并基于特征提取结果与预设画像生成模型，生成所述初始客户画像。S33: Perform feature extraction on the fused data, and generate the initial customer portrait based on the feature extraction result and a preset portrait generation model.

具体的，根据数据特征选择合适的相似度度量方法，如欧氏距离、余弦相似度、皮尔逊相关系数等，对特征标准数据和预设历史特征数据进行相似度计算，得到每对数据的相似度值。根据业务需求设定一个相似度阈值，提取相似度超过预设相似度阈值的历史特征数据，形成待融合数据集。Specifically, according to the data characteristics, select a suitable similarity measurement method, such as Euclidean distance, cosine similarity, Pearson correlation coefficient, etc., calculate the similarity of the feature standard data and the preset historical feature data, and obtain the similarity value of each pair of data. Set a similarity threshold according to business needs, extract historical feature data whose similarity exceeds the preset similarity threshold, and form a data set to be fused.

进一步的，根据数据的重要性或可靠性定义融合权重，可以设置预设权重或通过模型学习权重。对特征标准数据和待融合数据进行加权计算，生成融合数据。从融合数据中选择对客户画像生成有重要影响的关键特征，如客户年龄、性别、交易频率、产品偏好等。使用数据分析和统计方法(如相关分析、主成分分析等)分析选定特征的分布和关系，确保数据的准确性和有效性。基于特征提取结果，构建客户画像模型。模型可以采用规则引擎或机器学习算法(如决策树、随机森林等)来生成客户画像。根据画像模型生成初始客户画像，包括客户的基本信息、行为特征、关系特征、偏好特征等。Furthermore, the fusion weight is defined according to the importance or reliability of the data. The preset weight can be set or the weight can be learned through the model. The feature standard data and the data to be fused are weighted to generate fused data. From the fused data, key features that have an important impact on the generation of customer portraits are selected, such as customer age, gender, transaction frequency, product preferences, etc. The distribution and relationship of the selected features are analyzed using data analysis and statistical methods (such as correlation analysis, principal component analysis, etc.) to ensure the accuracy and validity of the data. Based on the feature extraction results, a customer portrait model is constructed. The model can use a rule engine or a machine learning algorithm (such as a decision tree, random forest, etc.) to generate a customer portrait. Generate an initial customer portrait based on the portrait model, including the customer's basic information, behavioral characteristics, relationship characteristics, preference characteristics, etc.

利用相似历史特征数据和当前特征数据的加权融合，提高了初始客户画像的准确性和可靠性。加权融合方法能够有效整合不同来源的数据，充分利用历史数据中的有用信息，提升数据的利用率和价值。通过特征提取技术，提取关键特征，有助于构建更加精准和全面的客户画像，为后续的精准营销和个性化服务提供数据支持。采用机器学习算法和画像生成模型，实现自动化生成初始客户画像，提高画像生成的效率和准确性，减少人工干预。The weighted fusion of similar historical feature data and current feature data improves the accuracy and reliability of the initial customer portrait. The weighted fusion method can effectively integrate data from different sources, make full use of useful information in historical data, and improve the utilization and value of data. Through feature extraction technology, key features are extracted to help build a more accurate and comprehensive customer portrait, providing data support for subsequent precision marketing and personalized services. Machine learning algorithms and portrait generation models are used to automatically generate initial customer portraits, improve the efficiency and accuracy of portrait generation, and reduce manual intervention.

请参阅图3，图3为本申请产品推荐方法第二实施例中又一子流程示意图。Please refer to FIG. 3 , which is a schematic diagram of another sub-process in the second embodiment of the product recommendation method of the present application.

如图3所示，在本实施例中，步骤S4包括：As shown in FIG3 , in this embodiment, step S4 includes:

S41：对所述初始客户画像进行识别，确定客户潜在需求数据；S41: Identify the initial customer portrait and determine the customer's potential demand data;

具体的，分析初始客户画像中的关键特征，包括基本信息、行为特征和偏好特征。基于画像中的特征数据，使用需求预测模型(如协同过滤、分类算法等)识别客户的潜在需求。根据需求预测模型的输出，生成客户潜在需求数据，记录客户可能感兴趣的产品或服务类型。将客户的潜在需求数据进行分类，形成结构化的需求特征数据(客户潜在需求数据)。Specifically, analyze the key features of the initial customer portrait, including basic information, behavioral characteristics, and preference characteristics. Based on the feature data in the portrait, use demand forecasting models (such as collaborative filtering, classification algorithms, etc.) to identify the customer's potential needs. Based on the output of the demand forecasting model, generate customer potential demand data and record the types of products or services that the customer may be interested in. Classify the customer's potential demand data to form structured demand feature data (customer potential demand data).

需要说明的是，需求预测模型为基于客户历史数据和特征数据，预测客户潜在需求的算法模型。It should be noted that the demand forecasting model is an algorithmic model that predicts customer potential demand based on customer historical data and feature data.

S42：获取所述客户互动数据，并对所述客户互动数据与所述客户潜在需求数据进行所述标准化处理，得到补充特征数据；S42: Acquire the customer interaction data, and perform the standardization process on the customer interaction data and the customer potential demand data to obtain supplementary feature data;

具体的，收集客户的在线互动数据与线下互动数据，并处理缺失值、重复值和异常值，确保数据的完整性和准确性，去除无效数据。将互动数据和需求数据进行归一化处理，确保数据的一致性和可比性。将标准化处理后的客户互动数据和需求数据进行整合，形成补充特征数据。Specifically, collect online and offline interaction data of customers, and process missing values, duplicate values and outliers to ensure the integrity and accuracy of the data and remove invalid data. Normalize the interaction data and demand data to ensure the consistency and comparability of the data. Integrate the standardized customer interaction data and demand data to form supplementary feature data.

S43：基于所述补充特征数据，调整所述预设画像生成模型的参数与权重；S43: adjusting the parameters and weights of the preset portrait generation model based on the supplementary feature data;

S44：根据调整后的所述预设画像生成模型、所述补充特征数据以及所述特征提取结果，得到所述目标客户画像。S44: Obtain the target customer portrait according to the adjusted preset portrait generation model, the supplementary feature data and the feature extraction result.

具体的，将补充特征数据分为训练集和验证集，用于模型调整和验证。使用训练集数据，调整画像生成模型的参数与权重，确保模型能够更好地适应新的补充特征数据。使用验证集数据，评估调整后的模型性能，确保模型的准确性和泛化能力。根据验证结果，反复调整模型参数与权重，优化模型性能，确保模型能够准确生成客户画像。Specifically, the supplementary feature data is divided into a training set and a validation set for model adjustment and verification. The training set data is used to adjust the parameters and weights of the portrait generation model to ensure that the model can better adapt to the new supplementary feature data. The validation set data is used to evaluate the performance of the adjusted model to ensure the accuracy and generalization ability of the model. Based on the verification results, the model parameters and weights are repeatedly adjusted to optimize the model performance and ensure that the model can accurately generate customer portraits.

进一步的，将调整后的画像生成模型、补充特征数据以及先前提取的关键特征数据进行融合，形成完整的输入数据集。使用调整后的画像生成模型对整合后的数据集进行预测，生成目标客户画像。目标客户画像应包含客户的最新特征、行为模式和需求特征。Furthermore, the adjusted profile generation model, supplementary feature data, and previously extracted key feature data are integrated to form a complete input data set. The adjusted profile generation model is used to predict the integrated data set to generate a target customer profile. The target customer profile should include the customer's latest features, behavior patterns, and demand characteristics.

可选的，通过客户经理反馈、客户行为观察等方式，验证目标客户画像的准确性，确保其能够准确反映客户的最新需求和行为。将生成的目标客户画像应用于客户需求预测、精准营销、个性化推荐等场景，提高业务决策的精准度和效率。Optionally, verify the accuracy of the target customer portrait through feedback from account managers and observation of customer behavior to ensure that it accurately reflects the customer's latest needs and behaviors. Apply the generated target customer portrait to scenarios such as customer demand forecasting, precision marketing, and personalized recommendations to improve the accuracy and efficiency of business decisions.

通过实时获取和整合客户互动数据，动态更新客户画像，确保画像的时效性和准确性。结合客户的潜在需求数据和互动数据，提升需求预测的准确性，为精准营销和个性化推荐提供坚实基础。通过不断调整和优化画像生成模型的参数与权重，提高模型的预测性能，确保生成的客户画像更加全面和准确。生成的目标客户画像能够准确反映客户的最新需求和行为，支持精准营销、客户服务等业务决策，提升客户满意度和营销成功率。By acquiring and integrating customer interaction data in real time, the customer portrait is dynamically updated to ensure the timeliness and accuracy of the portrait. By combining the customer's potential demand data and interaction data, the accuracy of demand forecasting is improved, providing a solid foundation for precision marketing and personalized recommendations. By continuously adjusting and optimizing the parameters and weights of the portrait generation model, the model's prediction performance is improved to ensure that the generated customer portrait is more comprehensive and accurate. The generated target customer portrait can accurately reflect the customer's latest needs and behaviors, support business decisions such as precision marketing and customer service, and improve customer satisfaction and marketing success rate.

本实施例通过先对预设产品数据与客户数据进行分类，得到多类型数据；接着对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。本实施例通过数据分类和标准化处理，生成初始客户画像，并动态更新客户画像以实时反映客户需求，最终基于目标客户画像精准推荐产品，提高了数据处理一致性和推荐精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果，提升了营销效能。This embodiment first classifies the preset product data and customer data to obtain multiple types of data; then standardizes the multiple types of data to obtain feature standard data; then generates an initial customer portrait based on the feature standard data and the preset historical feature data; then obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; finally, determines the target recommended product based on the target customer portrait. This embodiment generates an initial customer portrait through data classification and standardization, and dynamically updates the customer portrait to reflect customer needs in real time, and finally accurately recommends products based on the target customer portrait, thereby improving data processing consistency and recommendation accuracy, ensuring that the recommended products are highly matched with customer needs, improving the effect of product recommendations, and improving marketing efficiency.

基于上述第二实施例，提出本申请产品推荐方法的第三实施例。请参阅图4，图4为本申请产品推荐方法第三实施例中一子流程示意图。Based on the above second embodiment, a third embodiment of the product recommendation method of the present application is proposed. Please refer to Figure 4, which is a schematic diagram of a sub-flow in the third embodiment of the product recommendation method of the present application.

在本实施例中，步骤S41包括：In this embodiment, step S41 includes:

S411：基于预设特征识别算法，对所述初始客户画像进行特征识别，得到行为特征、关系特征以及偏好特征；S411: Based on a preset feature recognition algorithm, feature recognition is performed on the initial customer portrait to obtain behavior features, relationship features, and preference features;

具体的，准备初始客户画像数据，包括客户的基本信息、历史行为数据、互动记录等。选择或开发特征识别算法，算法可以是基于规则的系统或机器学习模型。提取客户的行为数据，如交易频率、购买周期、浏览记录等。基于算法识别客户的行为特征，如消费习惯、活跃度等。提取客户与其他实体的关系数据，如客户与客户经理的互动、客户与其他客户的关系网络等。识别客户的关系特征，如与客户经理的联系频率、与其他企业的交互关系等。提取客户的偏好数据，如产品选择、反馈记录等。识别客户的偏好特征，如对某类产品的偏好、常购买的产品类型等。Specifically, prepare initial customer portrait data, including basic customer information, historical behavior data, interaction records, etc. Select or develop feature recognition algorithms, which can be rule-based systems or machine learning models. Extract customer behavior data, such as transaction frequency, purchase cycle, browsing history, etc. Identify customer behavior characteristics based on algorithms, such as consumption habits, activity, etc. Extract relationship data between customers and other entities, such as interactions between customers and account managers, and relationship networks between customers and other customers. Identify customer relationship characteristics, such as frequency of contact with account managers, interactive relationships with other companies, etc. Extract customer preference data, such as product selection, feedback records, etc. Identify customer preference characteristics, such as preference for a certain type of product, types of products frequently purchased, etc.

S412：根据所述行为特征、所述关系特征以及所述偏好特征，对客户行为与客户需求进行预测，得到行为预测数据与需求预测数据；S412: predicting customer behavior and customer demand based on the behavior characteristics, the relationship characteristics, and the preference characteristics to obtain behavior prediction data and demand prediction data;

将识别得到的行为特征作为输入，构建行为预测模型，模型可以采用时间序列分析、回归分析、机器学习算法等。使用行为预测模型，对客户未来的行为进行预测，得到行为预测数据，如未来购买次数、未来交易金额、未来访问频率等。将识别得到的偏好特征和关系特征作为输入，构建需求预测模型，模型可以采用分类算法、关联规则挖掘、协同过滤等。使用需求预测模型，对客户的需求进行预测，得到需求预测数据，如可能感兴趣的产品、可能需要的服务、潜在的购买意图等。The identified behavioral features are used as input to build a behavioral prediction model. The model can use time series analysis, regression analysis, machine learning algorithms, etc. Use the behavioral prediction model to predict the customer's future behavior and obtain behavioral prediction data, such as the number of future purchases, future transaction amounts, and future visit frequencies. Use the identified preference features and relationship features as input to build a demand prediction model. The model can use classification algorithms, association rule mining, collaborative filtering, etc. Use the demand prediction model to predict customer demand and obtain demand prediction data, such as products that may be of interest, services that may be needed, and potential purchase intentions.

S413：对所述行为预测数据与所述需求预测数据进行需求概率评估，并将超过预设需求概率阈值的概率评估结果对应的所述行为预测数据或所述需求预测数据作为所述客户潜在需求数据。S413: Performing demand probability evaluation on the behavior prediction data and the demand prediction data, and taking the behavior prediction data or the demand prediction data corresponding to the probability evaluation result exceeding a preset demand probability threshold as the customer potential demand data.

具体的，将行为预测数据和需求预测数据作为输入，构建需求概率评估模型，模型可以采用概率论方法、贝叶斯网络、机器学习分类模型等。对每个预测数据进行概率评估，计算其发生的概率值。根据业务需求设定需求概率阈值，如0.7或0.8，作为判断客户需求的标准。将超过预设需求概率阈值的概率评估结果对应的行为预测数据和需求预测数据筛选出来，作为客户的潜在需求数据。将筛选后的数据整合，形成客户潜在需求数据集，记录客户可能感兴趣的产品或服务类型。Specifically, the behavior prediction data and demand prediction data are used as input to build a demand probability assessment model. The model can use probability theory methods, Bayesian networks, machine learning classification models, etc. Perform a probability assessment on each prediction data and calculate its probability value. Set a demand probability threshold according to business needs, such as 0.7 or 0.8, as a standard for judging customer demand. Filter out the behavior prediction data and demand prediction data corresponding to the probability assessment results that exceed the preset demand probability threshold as the customer's potential demand data. Integrate the filtered data to form a customer potential demand data set, recording the types of products or services that the customer may be interested in.

需要说明的是，特征识别算法为用于识别和提取数据中特定特征的算法，可以是基于规则的系统或机器学习模型。行为预测数据为通过预测模型对客户未来行为的预测结果，如未来购买次数、未来交易金额等。需求预测数据为通过预测模型对客户未来需求的预测结果，如可能感兴趣的产品、可能需要的服务等。需求概率评估指的是对预测数据进行概率评估，计算其发生的概率值。需求概率阈值是用于判断客户需求的标准概率值。It should be noted that feature recognition algorithms are algorithms used to identify and extract specific features from data, and can be rule-based systems or machine learning models. Behavior prediction data is the prediction result of customers' future behavior through prediction models, such as the number of future purchases, future transaction amounts, etc. Demand prediction data is the prediction result of customers' future needs through prediction models, such as products that may be of interest, services that may be needed, etc. Demand probability assessment refers to the probability assessment of prediction data and the calculation of its probability value. The demand probability threshold is the standard probability value used to judge customer demand.

通过特征识别和预测模型，精确预测客户的行为和需求，提高需求预测的准确性。通过行为特征、关系特征和偏好特征的识别与分析，全面了解客户需求，提供个性化推荐和服务。通过需求概率评估，筛选出高概率的潜在需求数据，优化营销策略，提升营销效果和客户满意度。基于客户互动数据和需求预测数据的实时更新，动态调整客户画像，确保画像的时效性和准确性。Through feature recognition and prediction models, we can accurately predict customer behavior and needs and improve the accuracy of demand forecasting. Through the identification and analysis of behavioral features, relationship features, and preference features, we can fully understand customer needs and provide personalized recommendations and services. Through demand probability assessment, we can screen out high-probability potential demand data, optimize marketing strategies, and improve marketing effectiveness and customer satisfaction. Based on the real-time update of customer interaction data and demand forecast data, we can dynamically adjust customer portraits to ensure the timeliness and accuracy of portraits.

请参阅图5，图5为本申请产品推荐方法第三实施例中又一子流程示意图。Please refer to FIG. 5 , which is a schematic diagram of another sub-process in the third embodiment of the product recommendation method of the present application.

在本实施例中，步骤S5包括：In this embodiment, step S5 includes:

S51：对所述目标客户画像进行关键特征提取，得到关键特征数据；S51: extract key features from the target customer portrait to obtain key feature data;

S52：将所述关键特征数据与所述预设产品数据进行归一化处理，得到归一化关键特征数据与归一化产品数据；S52: normalizing the key feature data and the preset product data to obtain normalized key feature data and normalized product data;

S53：将所述归一化关键特征数据与所述归一化产品数据中的各产品特征数据进行匹配；S53: Matching the normalized key feature data with each product feature data in the normalized product data;

S54：根据匹配度高于预设匹配度阈值对应的产品特征数据，确定所述目标推荐产品。S54: Determine the target recommended product according to the product feature data corresponding to the matching degree being higher than the preset matching degree threshold.

具体的，准备目标客户画像数据，包括客户的基本信息、行为特征、关系特征、偏好特征等。确定对产品推荐有显著影响的关键特征，如企业财务报表信息、供应链信息、数字化程度等。从目标客户画像中提取选定的关键特征，形成关键特征数据集。Specifically, prepare target customer portrait data, including basic customer information, behavioral characteristics, relationship characteristics, preference characteristics, etc. Determine key features that have a significant impact on product recommendations, such as corporate financial statement information, supply chain information, degree of digitization, etc. Extract selected key features from the target customer portrait to form a key feature data set.

进一步的，准备银行产品数据，包括产品名称、类别、功能、价格、适用客户群等信息。对关键特征数据和产品数据进行归一化处理，确保数据在统一尺度上便于比较。根据数据特性选择合适的匹配算法，如余弦相似度、欧氏距离、皮尔逊相关系数等。对归一化关键特征数据和归一化产品数据进行相似度计算，得到每个客户特征数据与每个产品特征数据之间的匹配度。根据业务需求设定一个匹配度阈值，如0.7或0.8，作为推荐标准。将匹配度高于预设阈值的产品特征数据筛选出来，确定为目标推荐产品。根据筛选结果生成推荐产品列表，包含所有符合匹配度要求的产品。Furthermore, prepare bank product data, including product name, category, function, price, applicable customer group and other information. Normalize key feature data and product data to ensure that the data is easy to compare on a unified scale. Select appropriate matching algorithms according to data characteristics, such as cosine similarity, Euclidean distance, Pearson correlation coefficient, etc. Calculate the similarity of normalized key feature data and normalized product data to obtain the matching degree between each customer feature data and each product feature data. Set a matching threshold, such as 0.7 or 0.8, as a recommendation standard based on business needs. Filter out product feature data with a matching degree higher than the preset threshold and determine them as target recommended products. Generate a recommended product list based on the screening results, including all products that meet the matching requirements.

通过提取关键特征数据，确保推荐产品与客户需求高度匹配，提高推荐的准确性。归一化处理保证了客户特征数据与产品数据的可比性，减少了数据异构性带来的影响。采用相似度匹配算法，提高了产品推荐的效率和准确性，确保推荐结果的可靠性。精准的产品推荐增强了客户的个性化体验，提高了客户满意度和忠诚度，提升了营销效果。By extracting key feature data, we ensure that the recommended products are highly matched with customer needs and improve the accuracy of recommendations. Normalization ensures the comparability of customer feature data and product data, and reduces the impact of data heterogeneity. The similarity matching algorithm is used to improve the efficiency and accuracy of product recommendations and ensure the reliability of recommendation results. Accurate product recommendations enhance customers' personalized experience, improve customer satisfaction and loyalty, and enhance marketing effectiveness.

基于上述第二实施例，在本实施例中，在步骤S5之后，还包括：Based on the above second embodiment, in this embodiment, after step S5, the following is further included:

S5a：获取购买详情信息，并判断所述购买详情信息中是否包括所述目标推荐产品；S5a: Acquire purchase details information, and determine whether the purchase details information includes the target recommended product;

具体的，从银行的交易系统、CRM系统等收集客户的购买详情信息，包括购买的产品名称、购买时间、购买金额等。对收集到的购买详情信息进行整理和存储，确保数据的完整性和准确性。将购买详情信息与目标推荐产品进行比对，判断客户是否购买了推荐的产品。记录比对结果，标记哪些推荐产品被客户购买，哪些未被购买。Specifically, collect the customer's purchase details from the bank's transaction system, CRM system, etc., including the name of the product purchased, the time of purchase, the amount of purchase, etc. Organize and store the collected purchase details to ensure the integrity and accuracy of the data. Compare the purchase details with the target recommended products to determine whether the customer has purchased the recommended products. Record the comparison results and mark which recommended products were purchased by the customer and which were not.

S5b：若不包括，获取客户反馈信息，并根据所述客户反馈信息，确定购买决策影响因素；S5b: If not included, obtain customer feedback information, and determine the factors affecting the purchase decision based on the customer feedback information;

具体的，通过多种渠道收集客户反馈信息，如客户调查问卷、电话回访、在线评价等。对收集到的客户反馈信息进行整理和分类，记录客户对推荐产品的意见和建议。对客户反馈信息进行分析，识别客户未购买推荐产品的原因，如价格过高、产品不符合需求、推荐时机不对等。基于反馈分析，确定影响客户购买决策的主要因素。Specifically, collect customer feedback information through various channels, such as customer questionnaires, telephone follow-up, online reviews, etc. Sort and classify the collected customer feedback information, and record customers' opinions and suggestions on recommended products. Analyze customer feedback information to identify the reasons why customers did not purchase recommended products, such as too high prices, products that do not meet needs, wrong timing of recommendations, etc. Based on feedback analysis, determine the main factors that affect customers' purchasing decisions.

S5c：基于所述购买决策影响因素与所述关键特征数据，生成产品推荐优化策略。S5c: Generate a product recommendation optimization strategy based on the purchase decision influencing factors and the key feature data.

具体的，将购买决策影响因素与客户的关键特征数据进行整合，形成综合数据集。分析关键特征数据与购买决策影响因素之间的关系，识别影响推荐效果的关键变量。基于综合数据集，调整产品推荐模型的参数和权重，优化推荐策略。根据识别出的关键变量和影响因素，调整推荐规则和逻辑，如调整推荐产品的价格范围、推荐时间、产品特性等。将优化后的推荐策略应用到产品推荐系统中，生成新的推荐方案。通过客户反馈和购买详情信息验证优化策略的效果，不断调整和完善推荐策略。Specifically, integrate the factors influencing the purchase decision with the key feature data of the customer to form a comprehensive data set. Analyze the relationship between the key feature data and the factors influencing the purchase decision, and identify the key variables that affect the recommendation effect. Based on the comprehensive data set, adjust the parameters and weights of the product recommendation model to optimize the recommendation strategy. According to the identified key variables and influencing factors, adjust the recommendation rules and logic, such as adjusting the price range, recommendation time, product features, etc. of the recommended products. Apply the optimized recommendation strategy to the product recommendation system to generate a new recommendation plan. Verify the effect of the optimization strategy through customer feedback and purchase details information, and continuously adjust and improve the recommendation strategy.

需要说明的是，购买详情信息为记录客户购买行为的详细数据，包括购买的产品名称、时间、金额等。客户反馈信息为客户对产品和服务的意见和建议，可以通过调查问卷、电话回访、在线评价等方式收集。购买决策影响因素指的是影响客户购买决策的主要原因，如价格、产品特性、推荐时机等。产品推荐优化策略是基于分析结果，调整推荐模型和规则，优化推荐效果的策略。It should be noted that purchase details are detailed data that record customer purchase behavior, including the name, time, and amount of the product purchased. Customer feedback information refers to customers' opinions and suggestions on products and services, which can be collected through questionnaires, telephone follow-up, online evaluation, etc. Purchase decision influencing factors refer to the main reasons that affect customers' purchase decisions, such as price, product features, and recommendation timing. Product recommendation optimization strategy is a strategy that adjusts recommendation models and rules based on analysis results to optimize recommendation effects.

通过分析未购买推荐产品的原因，针对性地优化推荐策略，提高推荐产品的匹配度和客户接受度。收集并分析客户反馈，及时调整推荐策略，提供更加符合客户需求的推荐方案，提升客户满意度。结合购买详情信息和客户反馈，不断优化推荐模型和规则，确保推荐系统能够实时适应市场和客户需求的变化，提高推荐效果。通过数据分析和反馈机制，基于真实数据驱动产品推荐优化，增强推荐策略的科学性和有效性，提高业务决策的精准度。By analyzing the reasons for not purchasing recommended products, we can optimize the recommendation strategy in a targeted manner to improve the matching degree and customer acceptance of recommended products. We collect and analyze customer feedback, adjust the recommendation strategy in a timely manner, provide recommendation solutions that better meet customer needs, and improve customer satisfaction. We combine purchase details and customer feedback to continuously optimize the recommendation model and rules to ensure that the recommendation system can adapt to changes in the market and customer needs in real time and improve the recommendation effect. Through data analysis and feedback mechanisms, we can drive product recommendation optimization based on real data, enhance the scientificity and effectiveness of recommendation strategies, and improve the accuracy of business decisions.

本实施例通过先对预设产品数据与客户数据进行分类，得到多类型数据；接着对多类型数据进行标准化处理，得到特征标准数据；然后基于特征标准数据与预设历史特征数据，生成初始客户画像；再获取客户互动数据，并基于客户互动数据与客户潜在需求数据更新初始客户画像，得到目标客户画像；最后根据目标客户画像，确定目标推荐产品。本实施例通过数据分类和标准化处理，生成初始客户画像，并动态更新客户画像以实时反映客户需求，最终基于目标客户画像精准推荐产品，提高了数据处理一致性和推荐精准度，确保推荐产品与客户需求高度匹配，提升了产品推荐的效果，提升了营销效能。This embodiment first classifies the preset product data and customer data to obtain multiple types of data; then standardizes the multiple types of data to obtain feature standard data; then generates an initial customer portrait based on the feature standard data and the preset historical feature data; then obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; finally, determines the target recommended product based on the target customer portrait. This embodiment generates an initial customer portrait through data classification and standardization, and dynamically updates the customer portrait to reflect customer needs in real time, and finally accurately recommends products based on the target customer portrait, thereby improving data processing consistency and recommendation accuracy, ensuring that the recommended products are highly matched with customer needs, improving the effect of product recommendations, and improving marketing efficiency.

示例性地，为了助于理解上述实施例的产品推荐方法的技术构思或技术原理，请参阅图6，图6为本申请产品推荐方法一实施例中产品推荐方法系统架构示例图。Exemplarily, in order to help understand the technical concept or technical principle of the product recommendation method of the above embodiment, please refer to Figure 6, which is an example diagram of the system architecture of the product recommendation method in an embodiment of the product recommendation method of this application.

如图6所示，在本实施例中，系统架构分为客户端和服务端两部分，通过Nginx进行请求分发，负载均衡到多个应用服务器。服务端包含多个模块，包括客户信息管理服务、客户画像特征服务和营销服务方案服务。客户终端包括客户的计算机和移动设备，通过互联网访问系统。通过Nginx进行请求分发，负载均衡到应用服务器。CRM服务网关负责接收和转发客户请求，进行权限控制和请求路由。客户信息管理服务提供客户信息查询和分析API。客户画像特征服务提供特征分析和客户画像生成。营销服务方案服务根据客户画像和需求，智能推荐产品和生成服务方案。应用服务器集群用于承载系统的各类应用服务。MySQL数据库用于存储客户信息、画像数据和营销方案，提供数据备份。RabbitMQ消息队列用于处理高并发请求，保障系统稳定性。Redis缓存用于缓存热点数据，减轻数据库压力。日志系统记录系统操作日志，便于监控和分析。ClickHouse和Gauss数据库用于高性能数据分析，存储客户画像和行为数据。文件服务器用于文件上传和下载，存储服务方案和相关文档。各模块之间通过服务调用进行数据同步，MySQL数据库定期备份数据。将来自不同来源的数据进行清洗、归一化处理，整合到特征信息数据库中。根据客户画像生成相应的产品推荐方案，并通过文件服务器进行存储和传输。该系统架构包括负载均衡、权限控制、高性能数据库、消息队列和缓存机制，提供一个高效、稳定的客户信息管理和产品推荐平台。通过特征数据的清洗和整合，精准生成客户画像，并基于画像智能推荐产品，优化产品推荐策略，提高客户满意度和产品推荐效果。As shown in FIG6 , in this embodiment, the system architecture is divided into two parts: the client and the server. Requests are distributed through Nginx and load balanced to multiple application servers. The server includes multiple modules, including customer information management service, customer portrait feature service and marketing service solution service. The customer terminal includes the customer's computer and mobile device, which access the system through the Internet. Requests are distributed through Nginx and load balanced to the application server. The CRM service gateway is responsible for receiving and forwarding customer requests, performing authority control and request routing. The customer information management service provides customer information query and analysis API. The customer portrait feature service provides feature analysis and customer portrait generation. The marketing service solution service intelligently recommends products and generates service solutions based on customer portraits and needs. The application server cluster is used to carry various application services of the system. The MySQL database is used to store customer information, portrait data and marketing solutions, and provide data backup. The RabbitMQ message queue is used to process high-concurrency requests and ensure system stability. The Redis cache is used to cache hot data and reduce database pressure. The log system records system operation logs for easy monitoring and analysis. ClickHouse and Gauss databases are used for high-performance data analysis and storage of customer portraits and behavior data. The file server is used for file upload and download, storage service solutions and related documents. Data synchronization is performed between modules through service calls, and the MySQL database regularly backs up data. Data from different sources is cleaned and normalized, and integrated into the feature information database. Corresponding product recommendation solutions are generated based on customer portraits, and stored and transmitted through file servers. The system architecture includes load balancing, permission control, high-performance databases, message queues and caching mechanisms, providing an efficient and stable customer information management and product recommendation platform. Through the cleaning and integration of feature data, customer portraits are accurately generated, and products are intelligently recommended based on the portraits, optimizing product recommendation strategies, and improving customer satisfaction and product recommendation effects.

进一步的，调用服务是指在系统各模块之间进行数据交换和功能调用的机制。在该系统架构中，通过服务网关和微服务架构实现各模块之间的调用和数据交互。各微服务通过服务网关进行注册与发现，确保服务的可访问性和动态调用。通过服务网关进行权限控制，确保只有授权的请求才能访问特定服务。服务网关和Nginx共同实现负载均衡，确保请求均匀分布到各应用服务器。重试机制指的是在系统调用过程中，如果某个服务出现临时故障，可以通过重试机制自动重新发起请求，尝试再次调用该服务。重试机制可提高系统的容错性，减少由于临时故障导致的请求失败。通过重试机制，可以在短暂的网络波动或服务不可用期间，确保请求能够成功。熔断机制用于保护系统免受大量失败请求的影响。当某个服务连续失败超过预设阈值时，系统会暂时停止对该服务的调用，防止系统资源被耗尽。防止大量失败请求消耗系统资源，确保其他服务的正常运行。通过熔断机制，可以在检测到服务恢复后，逐步恢复对该服务的调用，确保系统快速恢复正常。Furthermore, calling a service refers to a mechanism for exchanging data and calling functions between modules of the system. In this system architecture, calling and data interaction between modules are realized through the service gateway and microservice architecture. Each microservice is registered and discovered through the service gateway to ensure the accessibility and dynamic calling of the service. The service gateway is used to control permissions to ensure that only authorized requests can access specific services. The service gateway and Nginx jointly implement load balancing to ensure that requests are evenly distributed to each application server. The retry mechanism means that if a service has a temporary failure during the system call process, the retry mechanism can automatically re-initiate the request and try to call the service again. The retry mechanism can improve the fault tolerance of the system and reduce the request failure caused by temporary failures. The retry mechanism can ensure that the request can succeed during a short period of network fluctuations or service unavailability. The fuse mechanism is used to protect the system from a large number of failed requests. When a service fails continuously for more than a preset threshold, the system will temporarily stop calling the service to prevent system resources from being exhausted. Prevent a large number of failed requests from consuming system resources and ensure the normal operation of other services. Through the circuit breaker mechanism, after detecting that the service has been restored, calls to the service can be gradually restored to ensure that the system returns to normal quickly.

可选地，请参阅图7，图7为本申请产品推荐方法一实施例中数据处理示例图。Optionally, please refer to FIG. 7 , which is an example diagram of data processing in an embodiment of a product recommendation method of the present application.

如图7所示，收集多渠道数据后，整理各个场景下所需的客户及产品数据，进行分类(梳理营销相关系统)。数据清洗指的是处理缺失、重复、异常值，并进行归一化处理。统一数据格式(定义数据模板)指的是定义标签、关系、非结构化三类形式数据的属性模板。数据关联整合指的是统一客户ID和产品ID，将碎片化的数据整合到客户粒度和产品粒度的格式化表中。需要说明的是，数据来自不同的系统和场景，包括内部和外部数据源。经过多次处理，确保数据的质量和一致性，为后续的客户画像生成提供了可靠的数据基础。通过数据清洗和归一化处理，将各种形式的数据统一处理，便于后续分析和应用。As shown in Figure 7, after collecting multi-channel data, the customer and product data required for each scenario are sorted and classified (combing marketing-related systems). Data cleaning refers to processing missing, duplicate, and outlier values, and performing normalization. Unifying data formats (defining data templates) refers to defining attribute templates for three types of data: labels, relationships, and unstructured forms. Data association integration refers to unifying customer IDs and product IDs, and integrating fragmented data into formatted tables at customer and product granularity. It should be noted that the data comes from different systems and scenarios, including internal and external data sources. After multiple processing, the quality and consistency of the data are ensured, providing a reliable data foundation for the subsequent generation of customer portraits. Through data cleaning and normalization, various forms of data are processed in a unified manner to facilitate subsequent analysis and application.

本申请实施例还提供一种产品推荐装置，请参照图8，图8为本申请实施例产品推荐装置的模块结构示意图，所述产品推荐装置包括：The present application also provides a product recommendation device, please refer to FIG8 , which is a schematic diagram of the module structure of the product recommendation device of the present application embodiment, and the product recommendation device includes:

分类模块801，用于对预设产品数据与客户数据进行分类，得到多类型数据；The classification module 801 is used to classify the preset product data and customer data to obtain multiple types of data;

标准化模块802，用于对所述多类型数据进行标准化处理，得到特征标准数据；A standardization module 802 is used to perform standardization processing on the multi-type data to obtain characteristic standard data;

初始画像生成模块803，用于基于所述特征标准数据与预设历史特征数据，生成初始客户画像；An initial portrait generation module 803, used to generate an initial customer portrait based on the characteristic standard data and preset historical characteristic data;

画像更新模块804，用于获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据，更新所述初始客户画像，得到目标客户画像；A portrait updating module 804 is used to obtain customer interaction data, and based on the customer interaction data and customer potential demand data, update the initial customer portrait to obtain a target customer portrait;

产品推荐模块805，用于根据所述目标客户画像，确定目标推荐产品。The product recommendation module 805 is used to determine target recommended products based on the target customer portrait.

本申请实施例提供的产品推荐装置，采用上述实施例中的产品推荐方法，能够解决如何提高产品推荐的准确度，提升产品推荐效果的技术问题。与现有技术相比，本申请实施例提供的产品推荐装置的有益效果与上述实施例提供的产品推荐方法的有益效果相同，且所述产品推荐装置中的其他技术特征与上述实施例方法公开的特征相同，在此不做赘述。The product recommendation device provided in the embodiment of the present application adopts the product recommendation method in the above embodiment, which can solve the technical problem of how to improve the accuracy of product recommendation and enhance the product recommendation effect. Compared with the prior art, the beneficial effects of the product recommendation device provided in the embodiment of the present application are the same as the beneficial effects of the product recommendation method provided in the above embodiment, and the other technical features in the product recommendation device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

本申请提供一种产品推荐设备，产品推荐设备包括：至少一个处理器；以及，与至少一个处理器通信连接的存储器；其中，存储器存储有可被至少一个处理器执行的指令，指令被至少一个处理器执行，以使至少一个处理器能够执行上述实施例中的产品推荐方法。The present application provides a product recommendation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the product recommendation method in the above-mentioned embodiment.

下面参考图9，其示出了适于用来实现本申请实施例的产品推荐设备的结构示意图。本申请实施例中的产品推荐设备可以包括但不限于诸如移动电话、笔记本电脑、数字广播接收器、PDA(Personal Digital Assistant：个人数字助理)、PAD(PortableApplication Description：平板电脑)、PMP(Portable Media Player：便携式多媒体播放器)、车载终端(例如车载导航终端)等等的移动终端以及诸如数字TV、台式计算机等等的固定终端。图9示出的产品推荐设备仅仅是一个示例，不应对本申请实施例的功能和使用范围带来任何限制。Reference is made to Figure 9 below, which shows a schematic diagram of the structure of a product recommendation device suitable for implementing an embodiment of the present application. The product recommendation device in the embodiment of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. The product recommendation device shown in Figure 9 is only an example and should not bring any limitations to the functions and scope of use of the embodiments of the present application.

如图9所示，产品推荐设备可以包括处理装置1001(例如中央处理器、图形处理器等)，其可以根据存储在只读存储器(ROM：Read Only Memory)1002中的程序或者从存储装置1003加载到随机访问存储器(RAM：Random Access Memory)1004中的程序而执行各种适当的动作和处理。在RAM1004中，还存储有产品推荐设备操作所需的各种程序和数据。处理装置1001、ROM1002以及RAM1004通过总线1005彼此相连。输入/输出(I/O)接口1006也连接至总线。通常，以下系统可以连接至I/O接口1006：包括例如触摸屏、触摸板、键盘、鼠标、图像传感器、麦克风、加速度计、陀螺仪等的输入装置1007；包括例如液晶显示器(LCD：LiquidCrystal Display)、扬声器、振动器等的输出装置1008；包括例如磁带、硬盘等的存储装置1003；以及通信装置1009。通信装置1009可以允许产品推荐设备与其他设备进行无线或有线通信以交换数据。虽然图中示出了具有各种系统的产品推荐设备，但是应理解的是，并不要求实施或具备所有示出的系统。可以替代地实施或具备更多或更少的系统。As shown in FIG9 , the product recommendation device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the product recommendation device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input/output (I/O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I/O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the product recommendation device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a product recommendation device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

特别地，根据本申请公开的实施例，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，本申请公开的实施例包括一种计算机程序产品，其包括承载在计算机可读介质上的计算机程序，该计算机程序包含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装置从网络上被下载和安装，或者从存储装置1003被安装，或者从ROM1002被安装。在该计算机程序被处理装置1001执行时，执行本申请公开实施例的方法中限定的上述功能。In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a 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 a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

本申请提供的产品推荐设备，采用上述实施例中的产品推荐方法，能解决如何提高产品推荐的准确度，提升产品推荐效果的技术问题。与现有技术相比，本申请提供的产品推荐设备的有益效果与上述实施例提供的产品推荐方法的有益效果相同，且该产品推荐设备中的其他技术特征与上一实施例方法公开的特征相同，在此不做赘述。The product recommendation device provided by this application adopts the product recommendation method in the above embodiment, which can solve the technical problem of how to improve the accuracy of product recommendation and enhance the product recommendation effect. Compared with the prior art, the beneficial effects of the product recommendation device provided by this application are the same as the beneficial effects of the product recommendation method provided by the above embodiment, and the other technical features in the product recommendation device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

应当理解，本申请公开的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式的描述中，具体特征、结构、材料或者特点可以在任何的一个或多个实施例或示例中以合适的方式结合。It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以所述权利要求的保护范围为准。The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

本申请提供一种计算机可读存储介质，具有存储在其上的计算机可读程序指令(即计算机程序)，计算机可读程序指令用于执行上述实施例中的产品推荐方法。The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, wherein the computer-readable program instructions are used to execute the product recommendation method in the above-mentioned embodiment.

本申请提供的计算机可读存储介质例如可以是U盘，但不限于电、磁、光、电磁、红外线、或半导体的系统、系统或器件，或者任意以上的组合。计算机可读存储介质的更具体地例子可以包括但不限于：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器(RAM：Random Access Memory)、只读存储器(ROM：Read Only Memory)、可擦式可编程只读存储器(EPROM：Erasable Programmable Read Only Memory或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM：CD-Read Only Memory)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本实施例中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、系统或者器件使用或者与其结合使用。计算机可读存储介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：电线、光缆、RF(Radio Frequency：射频)等等，或者上述的任意合适的组合。The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

上述计算机可读存储介质可以是产品推荐设备中所包含的；也可以是单独存在，而未装配入产品推荐设备中。The computer-readable storage medium may be included in the product recommendation device; or may exist independently without being assembled into the product recommendation device.

上述计算机可读存储介质承载有一个或者多个程序，当上述一个或者多个程序被产品推荐设备执行时，使得产品推荐设备：对预设产品数据与客户数据进行分类，得到多类型数据；对所述多类型数据进行标准化处理，得到特征标准数据；基于所述特征标准数据与预设历史特征数据，生成初始客户画像；获取客户互动数据，并基于所述客户互动数据与客户潜在需求数据更新所述初始客户画像，得到目标客户画像；根据所述目标客户画像，确定目标推荐产品。可以以一种或多种程序设计语言或其组合来编写用于执行本申请的操作的计算机程序代码，上述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络——包括局域网(LAN：Local Area Network)或广域网(WAN：Wide Area Network)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the product recommendation device, the product recommendation device: classifies the preset product data and customer data to obtain multiple types of data; standardizes the multiple types of data to obtain feature standard data; generates an initial customer portrait based on the feature standard data and preset historical feature data; obtains customer interaction data, and updates the initial customer portrait based on the customer interaction data and customer potential demand data to obtain a target customer portrait; determines the target recommended product according to the target customer portrait. The computer program code for performing the operations of the present application can be written in one or more programming languages or a combination thereof, and the programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or it can be connected to an external computer (for example, through the Internet using an Internet service provider).

附图中的流程图和框图，图示了按照本申请各种实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，该模块、程序段、或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意，在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a sequence different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and/or flow chart, and the combination of the boxes in the block diagram and/or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

描述于本申请实施例中所涉及到的模块可以通过软件的方式实现，也可以通过硬件的方式来实现。其中，模块的名称在某种情况下并不构成对该单元本身的限定。The modules involved in the embodiments of the present application may be implemented by software or hardware, wherein the name of the module does not, in some cases, constitute a limitation on the unit itself.

本申请提供的可读存储介质为计算机可读存储介质，所述计算机可读存储介质存储有用于执行上述产品推荐方法的计算机可读程序指令(即计算机程序)，能够解决如何提高产品推荐的准确度，提升产品推荐效果的技术问题。与现有技术相比，本申请提供的计算机可读存储介质的有益效果与上述实施例提供的产品推荐方法的有益效果相同，在此不做赘述。The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned product recommendation method, and can solve the technical problem of how to improve the accuracy of product recommendations and enhance the product recommendation effect. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the product recommendation method provided in the above-mentioned embodiment, and will not be elaborated here.

本申请实施例提供一种计算机程序产品，包括计算机程序，所述计算机程序被处理器执行时实现如上述的产品推荐方法的步骤。An embodiment of the present application provides a computer program product, including a computer program, which implements the steps of the product recommendation method as described above when executed by a processor.

本申请提供的计算机程序产品能够解决如何提高产品推荐的准确度，提升产品推荐效果的技术问题。与现有技术相比，本申请实施例提供的计算机程序产品的有益效果与上述实施例提供的产品推荐方法的有益效果相同，在此不做赘述。The computer program product provided in this application can solve the technical problem of how to improve the accuracy of product recommendations and enhance the product recommendation effect. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiment of this application are the same as the beneficial effects of the product recommendation method provided in the above embodiment, and will not be repeated here.

以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利处理范围内。The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent processing scope of the present application.

Claims (10)

1. A method of product recommendation, comprising:

classifying preset product data and customer data to obtain multi-type data;

Carrying out standardization processing on the multi-type data to obtain characteristic standard data;

generating an initial customer portrait based on the feature standard data and preset historical feature data;

acquiring customer interaction data, and updating the initial customer portrait based on the customer interaction data and the customer potential demand data to obtain a target customer portrait;

and determining a target recommended product according to the target customer portrait.

2. The method of claim 1, wherein the step of normalizing the multi-type data to obtain feature standard data comprises:

performing data cleaning on the multi-type data to obtain cleaning data, wherein the data cleaning comprises one or more of missing value processing, repeated value processing and abnormal value processing;

Normalizing the cleaning data to obtain normalized data;

And integrating the normalized data into a preset formatting table to obtain the characteristic standard data.

3. The method of claim 1, wherein the step of generating an initial customer representation based on the feature criteria data and preset historical feature data comprises:

Extracting data, of which the similarity with the characteristic standard data exceeds a preset similarity threshold value, in the preset historical characteristic data to obtain data to be fused;

Carrying out weighted fusion on the data to be fused and the characteristic standard data to obtain fusion data;

And carrying out feature extraction on the fusion data, and generating the initial customer portrait based on a feature extraction result and a preset portrait generation model.

4. The method of claim 3, wherein the step of obtaining customer interaction data and updating the initial customer representation based on the customer interaction data and customer potential demand data to obtain a target customer representation comprises:

identifying the initial customer representation and determining the customer potential demand data;

acquiring the customer interaction data, and carrying out the standardized processing on the customer interaction data and the customer potential demand data to obtain supplementary feature data;

based on the supplementary feature data, adjusting parameters and weights of the preset portrait generation model;

And obtaining the target customer portrait according to the adjusted preset portrait generation model, the supplementary feature data and the feature extraction result.

5. The method of claim 4, wherein the step of identifying the initial customer representation and determining the customer potential demand data comprises:

Based on a preset feature recognition algorithm, performing feature recognition on the initial customer portrait to obtain behavior features, relationship features and preference features;

predicting the client behavior and the client demand according to the behavior characteristics, the relation characteristics and the preference characteristics to obtain behavior prediction data and demand prediction data;

And carrying out demand probability evaluation on the behavior prediction data and the demand prediction data, and taking the behavior prediction data or the demand prediction data corresponding to a probability evaluation result exceeding a preset demand probability threshold value as the potential demand data of the client.

6. The method of any one of claims 1 to 5, wherein the step of determining a target recommended product from the target customer representation comprises:

Extracting key features of the target customer portrait to obtain key feature data;

Normalizing the key characteristic data and the preset product data to obtain normalized key characteristic data and normalized product data;

matching the normalized key feature data with each product feature data in the normalized product data;

And determining the target recommended product according to the product characteristic data corresponding to the matching degree higher than the preset matching degree threshold.

7. The method of claim 6, further comprising, after the step of determining a target recommended product from the target customer representation:

acquiring purchase detail information, and judging whether the purchase detail information comprises the target recommended product or not;

If not, acquiring customer feedback information, and determining a purchase decision influencing factor according to the customer feedback information;

And generating a product recommendation optimization strategy based on the purchase decision influencing factors and the key feature data.

8. A product recommendation device, comprising:

the classification module is used for classifying preset product data and client data to obtain multi-type data;

The standardized module is used for carrying out standardized processing on the multi-type data to obtain characteristic standard data;

the initial portrait generation module is used for generating an initial customer portrait based on the feature standard data and the preset historical feature data;

the portrait updating module is used for acquiring customer interaction data, updating the initial customer portrait based on the customer interaction data and the customer potential demand data, and obtaining a target customer portrait;

And the product recommending module is used for determining a target recommended product according to the target customer portrait.

9. A computer device, the device comprising: a memory, a processor and a product recommendation program stored on the memory and executable on the processor, the product recommendation program being configured to implement the steps of the product recommendation method of any of claims 1 to 7.

10. A storage medium having stored thereon a product recommendation program which when executed by a processor implements the steps of the product recommendation method according to any one of claims 1 to 7.