WO2020206910A1

WO2020206910A1 - Product information pushing method, apparatus, computer device, and storage medium

Info

Publication number: WO2020206910A1
Application number: PCT/CN2019/102951
Authority: WO
Inventors: 张�杰
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-04-12
Filing date: 2019-08-28
Publication date: 2020-10-15
Also published as: CN110135888A; CN110135888B

Abstract

A product information pushing method, comprising: receiving a product information acquisition request, and extracting a product entity name from the product information acquisition request (S202); determining a word vector of the product entity name according to the product entity name and a preset word vector database, and generating a target vector of the product information acquisition request according to the word vector of the product entity name (S204); calculating a vector distance between the target vector and a vector of each triplet in a preset knowledge graph library, the triplets comprising two entity names and an entity relationship between the two entity names (S206); when a vector of a target triplet exists in the knowledge graph library, performing matching between the product entity name and the two entity names in the target triplet, determining an entity name matching the product entity name, determining corresponding product information according to an entity relationship in which the entity name matching the product entity name is located, and pushing the product information to a sender of the product information acquisition request, wherein the vector of the target triplet is a vector the vector distance of which from the target vector falls within a preset vector distance threshold range (S208); when the vector of the target triplet does not exist in the knowledge graph library, generating a service prompt according to the product information acquisition request, sending the service prompt to a customer service terminal, and pushing product information fed back by the customer service terminal to the sender of the product information acquisition request (S210).

Description

产品信息推送方法、装置、计算机设备和存储介质Product information push method, device, computer equipment and storage medium

相关申请的交叉引用Cross references to related applications

本申请要求于2019年4月12日提交中国专利局，申请号为2019102956934，申请名称为“产品信息推送方法、装置、计算机设备和存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on April 12, 2019. The application number is 2019102956934 and the application name is "Product Information Push Method, Apparatus, Computer Equipment, and Storage Medium". The entire content is by reference Incorporated in this application.

技术领域Technical field

本申请涉及一种产品信息推送方法、装置、计算机设备和存储介质。This application relates to a product information push method, device, computer equipment and storage medium.

背景技术Background technique

随着计算机技术的发展，出现了产品信息推送***。传统的产品信息推送***，是根据产品预先设想用户可能提出的问题，并给出对应问题答案，然后将问题以及对应的问题答案(通常包含产品信息)存储在产品信息数据库。在需要对产品在进行产品信息推送时，在接收到用户的产品信息获取请求后，通过提取产品信息获取请求中的关键字，基于关键字在产品信息数据库中对存储的问题进行全面检索和匹配，进而获得与匹配的问题对应的问题答案，根据该问题答案推送对应的产品信息。With the development of computer technology, product information push systems have emerged. The traditional product information push system pre-imagines the questions that users may ask based on the product, and gives the answers to the corresponding questions, and then stores the questions and the corresponding answer (usually including product information) in the product information database. When it is necessary to push the product information of the product, after receiving the user's product information acquisition request, the keyword in the product information acquisition request is extracted, and the stored questions in the product information database are fully retrieved and matched based on the keyword , And then obtain the question answer corresponding to the matched question, and push the corresponding product information according to the question answer.

然而，发明人意识到，目前的产品信息推送方式，需要基于关键字在数据库中进行全面搜索，效率低下，且在产品信息推送***未能存储有对应问题的情况下，将无法检索到到对应的问题及问题答案，影响产品信息推送的准确度。However, the inventor realizes that the current product information push method requires a comprehensive search in the database based on keywords, which is inefficient, and if the product information push system fails to store the corresponding problem, it will not be able to retrieve the corresponding The questions and answers to the questions affect the accuracy of product information push.

发明内容Summary of the invention

根据本申请公开的各种实施例，提供一种产品信息推送方法、装置、计算机设备和存储介质。According to various embodiments disclosed in the present application, a product information push method, device, computer equipment, and storage medium are provided.

一种产品信息推送方法包括：A method for pushing product information includes:

接收产品信息获取请求，从产品信息获取请求中提取产品实体名称；Receive the product information acquisition request, and extract the product entity name from the product information acquisition request;

根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据产品实体名称的词向量，生成产品信息获取请求的目标向量；Determine the word vector of the product entity name according to the product entity name and the preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

计算目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，三元组包括两个实体名称以及两个实体名称之间的实体关系；Calculate the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library. The triplet includes two entity names and the entity relationship between the two entity names;

当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方，目标三元组的向量为与目标向量的向量距离在预设的向量距离阈值范围内的向量；及When there is a vector of target triples in the knowledge graph database, the product entity name is matched with the two entity names in the target triple, and the entity name that matches the product entity name is determined, and based on the matching product entity name The entity relationship where the entity name is located, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is the vector distance from the target vector within the preset vector distance threshold range Vector; and

当知识图谱库中不存在目标三元组的向量时，根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方。When the target triple vector does not exist in the knowledge graph database, a service prompt is generated according to the product information acquisition request, and the service prompt is sent to the customer service terminal, and the product information fed back by the customer service terminal is pushed to the sender of the product information acquisition request.

一种产品信息推送装置包括：A product information push device includes:

实体识别模块，用于接收产品信息获取请求，从产品信息获取请求中提取产品实体名称；The entity recognition module is used to receive the product information acquisition request, and extract the product entity name from the product information acquisition request;

第一处理模块，用于根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据产品实体名称的词向量，生成产品信息获取请求的目标向量；The first processing module is used to determine the word vector of the product entity name according to the product entity name and the preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

第二处理模块，用于计算目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，三元组包括两个实体名称以及两个实体名称之间的实体关系；The second processing module is used to calculate the vector distance between the target vector and the vector of each triad in the preset knowledge graph library, the triad includes two entity names and the entity relationship between the two entity names;

第一推送模块，用于当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方，目标三元组的向量为与目标向量的向量距离在预设的向量距离阈值范围内的向量；及The first push module is used to match the product entity name with two entity names in the target triple when there is a vector of the target triple in the knowledge graph database, determine the entity name that matches the product entity name, and According to the entity relationship of the entity name that matches the product entity name, determine the corresponding product information, and push the product information to the sender of the product information acquisition request. The vector of the target triplet is the vector distance from the target vector in the preset The vector within the distance threshold; and

第二推送模块，用于当知识图谱库中不存在目标三元组的向量时，根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方。The second push module is used to generate a service prompt according to the product information acquisition request when there is no target triple vector in the knowledge graph database, and send the service prompt to the customer service terminal, and push the product information fed back by the customer service terminal to the product The sender of the information acquisition request.

一种计算机设备，包括存储器和一个或多个处理器，所述存储器中储存有计算机可读指令，所述计算机可读指令被所述处理器执行时，使得所述一个或多个处理器执行以下步骤：A computer device, including a memory and one or more processors, the memory stores computer readable instructions, when the computer readable instructions are executed by the processor, the one or more processors execute The following steps:

一个或多个存储有计算机可读指令的非易失性计算机可读存储介质，计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器执行以下步骤：One or more non-volatile computer-readable storage media storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the following steps:

本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。The details of one or more embodiments of the application are set forth in the following drawings and description. Other features and advantages of this application will become apparent from the description, drawings and claims.

附图说明Description of the drawings

为了更清楚地说明本申请实施例中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。In order to more clearly describe the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

图1为根据一个或多个实施例中产品信息推送方法的应用场景图。Fig. 1 is an application scenario diagram of a product information push method according to one or more embodiments.

图2为根据一个或多个实施例中产品信息推送方法的流程示意图。Fig. 2 is a schematic flowchart of a method for pushing product information according to one or more embodiments.

图3为根据一个或多个实施例中图2中步骤S202的子流程示意图。Fig. 3 is a schematic diagram of a sub-flow of step S202 in Fig. 2 according to one or more embodiments.

图4为根据一个或多个实施例中图2中步骤S204的子流程示意图。Fig. 4 is a schematic diagram of the sub-flow of step S204 in Fig. 2 according to one or more embodiments.

图5为根据一个或多个实施例中图2中步骤S206的子流程示意图。Fig. 5 is a schematic diagram of a sub-flow of step S206 in Fig. 2 according to one or more embodiments.

图6为另一个实施例中产品信息推送方法的流程示意图；FIG. 6 is a schematic flowchart of a method for pushing product information in another embodiment;

图7为又一个实施例中产品信息推送方法的流程示意图；FIG. 7 is a schematic flowchart of a method for pushing product information in another embodiment;

图8为根据一个或多个实施例中产品信息推送装置的框图。Fig. 8 is a block diagram of a product information pushing device according to one or more embodiments.

图9为根据一个或多个实施例中计算机设备的框图。Figure 9 is a block diagram of a computer device according to one or more embodiments.

具体实施方式detailed description

为了使本申请的技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。In order to make the technical solutions and advantages of the present application clearer, the following further describes the present application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the application, but not to limit the application.

本申请提供的产品信息推送方法，可以应用于如图1所示的应用环境中。终端102与服务器104通过网络进行通信。服务器104接收终端102发送的产品信息获取请求，从产品信息获取请求中提取产品实体名称，根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据产品实体名称的词向量，生成产品信息获取请求的目标向量，计算目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，三元组包括两个实体名称以及两个实体名称之间的实体关系，当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方的终端102，目标三元组的向量为与目标向量的向量距离在预设的向量距离阈值范围内的向量，当知识图谱库中不存在目标三元组的向量时，根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方的终端102。终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备，服务器104可以用独立的服务器或者是多个服务器组成的服务器集群来实现。The product information push method provided in this application can be applied to the application environment shown in FIG. 1. The terminal 102 and the server 104 communicate through the network. The server 104 receives the product information acquisition request sent by the terminal 102, extracts the product entity name from the product information acquisition request, determines the word vector of the product entity name according to the product entity name and the preset word vector database, and according to the word vector of the product entity name , Generate the target vector of the product information acquisition request, and calculate the vector distance between the target vector and the vector of each triple in the preset knowledge graph library. The triple includes two entity names and the entity between the two entity names Relationship, when there is a vector of target triples in the knowledge graph database, the product entity name is matched with the two entity names in the target triple, and the entity name matching the product entity name is determined, and based on the product entity name The entity relationship where the matched entity name is located, the corresponding product information is determined, and the product information is pushed to the terminal 102 of the sender of the product information acquisition request. The vector of the target triplet is the vector distance from the target vector in the preset vector For vectors within the distance threshold, when there is no target triple vector in the knowledge graph database, a service prompt is generated according to the product information acquisition request, and the service prompt is sent to the customer service terminal, and the product information feedback from the customer service terminal is pushed to the product The terminal 102 of the sender of the information acquisition request. The terminal 102 may be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server 104 may be implemented by an independent server or a server cluster composed of multiple servers.

在其中一个实施例中，如图2所示，提供了一种产品信息推送方法，以该方法应用于图1中的服务器为例进行说明，包括以下步骤：In one of the embodiments, as shown in FIG. 2, a method for pushing product information is provided. Taking the method applied to the server in FIG. 1 as an example, the method includes the following steps:

S202：接收产品信息获取请求，从产品信息获取请求中提取产品实体名称。S202: Receive the product information acquisition request, and extract the product entity name from the product information acquisition request.

产品信息获取请求中包括与用户想要获取的产品信息对应的产品实体名称以及产品属性，产品实体名称指的是产品名称，产品属性包括容量以及颜色等。服务器接收产品信息获取请求，确定产品信息获取请求的数据类型，数据类型包括语音数据和文字数据，当数据类型为语音数据时，将语音数据转换为文字数据，当数据类型为文字数据时，不需要进行转换，对文字数据进行句法分析，确定文字数据的句法结构，将文字数据拆分为多个词语，确定拆分后的多个词语的词性，根据句法结构以及拆分后的多个词语的词性确定产品实体名称。The product information acquisition request includes the product entity name and product attributes corresponding to the product information that the user wants to acquire. The product entity name refers to the product name, and the product attributes include capacity and color. The server receives the product information acquisition request and determines the data type of the product information acquisition request. The data type includes voice data and text data. When the data type is voice data, the voice data is converted into text data. When the data type is text data, it is not Need to convert, perform syntactic analysis on text data, determine the syntactic structure of text data, split text data into multiple words, determine the part of speech of multiple words after splitting, according to the syntactic structure and multiple words after splitting The part of speech determines the product entity name.

S204：根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据产品实体名称的词向量，生成产品信息获取请求的目标向量。S204: Determine the word vector of the product entity name according to the product entity name and the preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name.

服务器根据产品实体名称遍历预设的词向量数据库，确定产品实体名称的词向量，根据预设的产品属性词库提取产品信息获取请求中的多个词语，根据预设的词向量数据库，确定提取出的各词语的词向量，根据提取出的各词语的词向量以及产品实体名称的词向量，生成产品信息获取请求的目标向量。在词向量数据库中存储了各个词语以及与各词语对应的词向量，产品属性词库中存储了用于表示产品属性的词语，词向量数据库和产品属性词库都可按照需要自行设置。The server traverses the preset word vector database according to the product entity name, determines the word vector of the product entity name, extracts multiple words in the product information acquisition request according to the preset product attribute word database, and determines the extraction according to the preset word vector database Based on the extracted word vectors of each word and the word vector of the product entity name extracted, the target vector of the product information acquisition request is generated. Each word and the word vector corresponding to each word are stored in the word vector database, and the product attribute word database stores words used to represent product attributes. Both the word vector database and the product attribute word database can be set as needed.

S206：计算目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，三元组包括两个实体名称以及两个实体名称之间的实体关系。S206: Calculate the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library, the triplet includes two entity names and an entity relationship between the two entity names.

知识图谱是一种数据结构，把世界上的知识组织成实体与实体之间的关系，实体与实体之间通过实体关系联系在一起。三元组指的是知识图谱中的两个实体以及两个实体之间的实体关系的组合。在本实施例中，预设的知识图谱库中的各三元组中包括两个实体名称以及两个实体名称之间的实体关系，两个实体名称分别与产品实体名称以及产品信息对应，实体关系与产品属性相对应。服务器获取预设的知识图谱库中各三元组的向量，各三元组的向量包括实体名称向量以及实体关系向量，获取目标向量中的产品实体名称向量以及产品属性向量，计算产品实体名称向量与各三元组的向量中的实体名称向量之间的第一词向量距离，计算产品属性向量与各三元组的向量中的实体关系向量之间的第二词向量距离，根据第一词向量距离和第二词向量距离，计算目标向量与各三元组的向量之间的向量距离。The knowledge graph is a data structure that organizes the world's knowledge into the relationship between entities and the entities are linked together through the entity relationship. A triple refers to the combination of two entities in the knowledge graph and the entity relationship between the two entities. In this embodiment, each triplet in the preset knowledge graph database includes two entity names and the entity relationship between the two entity names. The two entity names correspond to the product entity name and product information, respectively. Relationships correspond to product attributes. The server obtains the vector of each triplet in the preset knowledge graph library. The vector of each triple includes the entity name vector and the entity relationship vector, obtains the product entity name vector and the product attribute vector in the target vector, and calculates the product entity name vector The first word vector distance between the entity name vector in the vector of each triplet and the second word vector distance between the product attribute vector and the entity relationship vector in the vector of each triplet is calculated according to the first word The vector distance and the second word vector distance are used to calculate the vector distance between the target vector and the vector of each triple.

S208：当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方，目标三元组的向量为与目标向量的向量距离在预设的向量距离阈值范围内的向量。S208: When there is a vector of target triples in the knowledge graph database, the product entity name is matched with the two entity names in the target triple, and the entity name matching the product entity name is determined, and based on the product entity name The entity relationship of the matching entity name is determined, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is the vector distance from the target vector within the preset vector distance threshold range Vector within.

向量距离阈值可按照需要自行设置。当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，在目标三元组中包括两个实体名称，其中一个实体名称与产品实体名称相对应，另一个实体名称与产品信息相对应，并根据与产品实体名称匹配的实体名称所在的实体关系以及目标三元组的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方，目标三元组的实体关系具体可以为产品实体名称与产品信息之间的对应关系。进一步的，实体关系可以为产品属性。The vector distance threshold can be set as needed. When there is a vector of target triples in the knowledge graph database, the product entity name is matched with the two entity names in the target triple, and the entity name that matches the product entity name is determined. The target triple includes two Entity names, one entity name corresponds to the product entity name, the other entity name corresponds to the product information, and the corresponding entity name is determined based on the entity relationship of the entity name matching the product entity name and the entity relationship of the target triple And push the product information to the sender of the product information acquisition request. The entity relationship of the target triplet may specifically be the corresponding relationship between the product entity name and the product information. Further, the entity relationship may be a product attribute.

S210：当知识图谱库中不存在目标三元组的向量时，根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方。S210: When the target triple vector does not exist in the knowledge graph database, generate a service reminder according to the product information acquisition request, and send the service reminder to the customer service terminal, and push the product information feedback from the customer service terminal to the sending of the product information acquisition request square.

当知识图谱库中不存在目标三元组的向量时，就无法利用知识图谱库实现对产品信息获取请求的发送方的准确反馈，此时服务器可根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方，服务提示用于提示使用客服终端的客服人员反馈产品信息。When the target triple vector does not exist in the knowledge graph library, the knowledge graph library cannot be used to achieve accurate feedback on the sender of the product information acquisition request. At this time, the server can generate service prompts according to the product information acquisition request and provide services The reminder is sent to the customer service terminal, and the product information fed back by the customer service terminal is pushed to the sender of the product information acquisition request. The service reminder is used to prompt the customer service staff using the customer service terminal to feedback product information.

上述产品信息推送方法，通过预设知识图谱库存储三元组的向量，在接收到产品信息获取请求时，将产品信息获取请求表述为对应的目标向量，并通过目标向量与各三元组的向量之间的向量距离匹配出对应的目标三元组，当匹配结果为存在目标三元组时，可基于目标三元组确定对应的产品信息并进行推送，当匹配结果为不存在目标三元组时，通过向客服终端发送服务提示，可以从客服终端获得反馈的产品信息并进行推送。这种推送方式，通过存储三元组的向量的知识图谱，并用向量匹配的方式匹配出对应的产品信息进行推送，无需用关键字进行全面检索，提高了产品信息推送的效率，在知识图谱中没有匹配的目标三元组的向量的情况下，还从客服终端获得反馈的产品信息并进行推送，进一步提高了产品信息推送的准确性。In the above product information push method, the vector of the triples is stored through the preset knowledge graph library, and when the request for product information acquisition is received, the product information acquisition request is expressed as the corresponding target vector, and the target vector and each triplet The vector distance between the vectors matches the corresponding target triplet. When the matching result is that the target triplet exists, the corresponding product information can be determined and pushed based on the target triplet. When the matching result is that there is no target triplet When grouping, by sending service reminders to the customer service terminal, the feedback product information can be obtained from the customer service terminal and pushed. This kind of push method, by storing the knowledge map of the vector of the triples, and matching the corresponding product information by vector matching, it does not need to use keywords for comprehensive retrieval, which improves the efficiency of product information pushing. In the knowledge map When there is no matching target triple vector, the feedback product information is also obtained from the customer service terminal and pushed, which further improves the accuracy of product information push.

在其中一个实施例中，如图3所示，S202包括：In one of the embodiments, as shown in FIG. 3, S202 includes:

S302：接收产品信息获取请求，将产品信息获取请求转换为文字数据；S302: Receive the product information acquisition request, and convert the product information acquisition request into text data;

S304：对文字数据进行句法分析，确定文字数据的句法结构；S304: Perform syntactic analysis on the text data to determine the syntactic structure of the text data;

S306：将文字数据拆分为多个词语，确定拆分后的多个词语的词性；S306: Split the text data into multiple words, and determine the parts of speech of the multiple words after the split;

S308：根据句法结构以及拆分后的多个词语的词性确定产品实体名称。S308: Determine the product entity name according to the syntactic structure and the part of speech of multiple words after splitting.

服务器接收产品信息获取请求，确定产品信息获取请求的数据类型，数据类型包括语音数据和文字数据，当数据类型为语音数据时，将语音数据转换为文字数据，当数据类型为文字数据时，不需要进行转换，对文字数据进行句法分析，确定文字数据的句法结构，将文字数据拆分为多个词语，确定拆分后的多个词语的词性，根据句法结构以及拆分后的多个词语的词性确定产品实体名称。其中，确定文字数据的句法结构包括确定句子中的主语、谓语以及宾语等，词语的词性包括动词、形容词以及名词等，对文字数据进行筛选，筛选掉其中的谓语部分、动词以及形容词，根据筛选后的文字数据确定产品实体名称。The server receives the product information acquisition request and determines the data type of the product information acquisition request. The data type includes voice data and text data. When the data type is voice data, the voice data is converted into text data. When the data type is text data, it is not Need to convert, perform syntactic analysis on text data, determine the syntactic structure of text data, split text data into multiple words, determine the part of speech of multiple words after splitting, according to the syntactic structure and multiple words after splitting The part of speech determines the product entity name. Among them, determining the syntactic structure of the text data includes determining the subject, predicate, and object in the sentence. The part of speech of the word includes verbs, adjectives, and nouns. The text data is filtered, and the predicate part, verbs and adjectives are filtered out. The following text data determines the product entity name.

上述实施例，对产品信息获取请求的文字数据进行句法分析和拆分，根据句法结构和拆分后的多个词语的词性，确定产品实体名称，实现了对产品实体名称的获取。In the above embodiment, the text data of the product information acquisition request is syntactically analyzed and split, and the product entity name is determined according to the syntax structure and the part of speech of multiple words after splitting, thereby achieving the acquisition of the product entity name.

在其中一个实施例中，S308包括：In one of the embodiments, S308 includes:

S310：根据句法结构确定各句子中的主语、谓语以及宾语；S310: Determine the subject, predicate, and object in each sentence according to the syntactic structure;

S312：根据句法结构以及拆分后的多个词语的词性筛选文字数据，筛选掉文字数据中的谓语、动词以及形容词，词性包括动词、形容词以及名词；S312: Filter the text data according to the syntactic structure and the parts of speech of the split words, and filter out the predicates, verbs and adjectives in the text data. The parts of speech include verbs, adjectives and nouns;

S314：根据筛选后的文字数据确定产品实体名称。S314: Determine the product entity name according to the filtered text data.

服务器首先根据句法结构确定各句子中的主语、谓语以及宾语，并根据句法结构对各句子进行标注，根据各词语的词性对各词语进行标注，然后根据标注对文字数据进行筛选，筛选掉文字数据中的谓语、动词以及形容词，最后统计筛选后的文字数据中各词语出现的次数，根据各词语出现的次数确定产品实体名称。具体的，将出现次数大于预设的次数阈值的名词作为产品实体名称。The server first determines the subject, predicate, and object of each sentence according to the syntactic structure, annotates each sentence according to the syntactic structure, annotates each word according to the part of speech of each word, and then filters the text data according to the annotation to filter out the text data The predicates, verbs and adjectives in, finally count the number of occurrences of each word in the filtered text data, and determine the product entity name according to the number of occurrences of each word. Specifically, nouns whose appearance times are greater than a preset frequency threshold are used as product entity names.

在其中一个实施例中，如图4所示，S204包括：In one of the embodiments, as shown in FIG. 4, S204 includes:

S402：根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量；S402: Determine the word vector of the product entity name according to the product entity name and the preset word vector database;

S404：根据预设的产品属性词库提取产品信息获取请求中的多个词语，根据预设的词向量数据库，确定提取出的各词语的词向量；S404: Extract multiple words in the product information acquisition request according to the preset product attribute word database, and determine the word vector of each word extracted according to the preset word vector database;

S406：根据提取出的各词语的词向量以及产品实体名称的词向量，生成产品信息获取请求的目标向量。S406: Generate a target vector of the product information acquisition request according to the word vector of each word extracted and the word vector of the product entity name.

服务器根据预设的词向量数据库，确定产品实体名称的词向量，将产品信息获取请求转换为文字数据并进行拆分，提取产品信息获取请求中的多个词语，进而根据预设的词向量数据库，确定提取出的各词语的词向量，根据提取出的各词语的词向量以及产品实体名称的词向量，生成产品信息获取请求的目标向量。预设的词向量数据库中包括各产品实体名称以及各产品属性的词向量，常见的产品属性包括容量、颜色等。提取产品信息获取请求中的多个词语，实际是在提取产品信息获取请求中的能表示产品属性的词语，举例说明，可以根据预设的产品属性词库遍历已拆分的产品信息获取请求的文字数据，确定文字数据中能表示产品属性的词语。The server determines the word vector of the product entity name according to the preset word vector database, converts the product information acquisition request into text data and splits it, extracts multiple words in the product information acquisition request, and then according to the preset word vector database , Determine the word vector of each word extracted, and generate the target vector of the product information acquisition request according to the word vector of each word extracted and the word vector of the product entity name. The preset word vector database includes the name of each product entity and the word vector of each product attribute. Common product attributes include capacity, color, etc. Extracting multiple words in the product information acquisition request is actually extracting the words in the product information acquisition request that can represent product attributes. For example, you can traverse the split product information acquisition request according to the preset product attribute thesaurus Text data, to determine the words in the text data that can express the attributes of the product.

上述实施例，根据预设的词向量数据库，确定产品实体名称的词向量，提取产品信息获取请求中的多个词语，根据预设的词向量数据库，确定提取出的各词语的词向量，根据提取出的各词语的词向量以及产品实体名称的词向量，生成产品信息获取请求的目标向量，实现了对目标向量的获取。In the above-mentioned embodiment, the word vector of the product entity name is determined according to the preset word vector database, multiple words in the product information acquisition request are extracted, and the word vector of each word extracted is determined according to the preset word vector database. The word vector of each word extracted and the word vector of the product entity name are generated to generate the target vector of the product information acquisition request, which realizes the acquisition of the target vector.

在其中一个实施例中，如图5所示，S206包括：In one of the embodiments, as shown in FIG. 5, S206 includes:

S502：获取预设的知识图谱库中各三元组的向量中的实体名称向量以及实体关系向量；S502: Obtain the entity name vector and the entity relationship vector in the vector of each triplet in the preset knowledge graph library;

S504：获取目标向量中的产品实体名称向量以及产品属性向量；S504: Obtain the product entity name vector and the product attribute vector in the target vector;

S506：计算产品实体名称向量与各三元组的向量中的实体名称向量之间的第一词向量距离；S506: Calculate the first word vector distance between the product entity name vector and the entity name vector in each triplet vector;

S508：计算产品属性向量与各三元组的向量中的实体关系向量之间的第二词向量距离；S508: Calculate the second word vector distance between the product attribute vector and the entity relationship vector in the vector of each triplet;

S510：根据第一词向量距离和第二词向量距离，计算目标向量与各三元组的向量之间的向量距离。S510: Calculate the vector distance between the target vector and the vector of each triple according to the first word vector distance and the second word vector distance.

服务器获取预设的知识图谱库中各三元组的向量中的实体名称向量以及实体关系向量，获取目标向量中的产品实体名称向量以及产品属性向量，计算产品实体名称向量与各三元组的向量中的实体名称向量之间的第一词向量距离，计算产品属性向量与各三元组的向量中的实体关系向量之间的第二词向量距离，根据第一词向量距离和第二词向量距离，计算目标向量与各三元组的向量之间的向量距离。目标向量与各三元组的向量之间的向量距离可以为第一词向量距离和第二词向量距离之和。The server obtains the entity name vector and the entity relationship vector in the vector of each triple in the preset knowledge graph library, obtains the product entity name vector and the product attribute vector in the target vector, and calculates the product entity name vector and each triple The first word vector distance between the entity name vectors in the vector, calculate the second word vector distance between the product attribute vector and the entity relationship vector in each triplet vector, according to the first word vector distance and the second word Vector distance, calculate the vector distance between the target vector and the vector of each triple. The vector distance between the target vector and the vector of each triplet may be the sum of the first word vector distance and the second word vector distance.

上述实施例，获取预设的知识图谱库中各三元组的向量中的实体名称向量以及实体关系向量，获取目标向量中的产品实体名称向量以及产品属性向量，计算产品实体名称向量与各三元组的向量中的实体名称向量之间的第一词向量距离，计算产品属性向量与各三元组的向量中的实体关系向量之间的第二词向量距离，根据第一词向量距离和第二词向量距离，计算目标向量与各三元组的向量之间的向量距离，实现了目标向量与各三元组的向量之间的向量距离的确定。In the above embodiment, the entity name vector and the entity relationship vector in the vector of each triple in the preset knowledge graph library are obtained, the product entity name vector and the product attribute vector in the target vector are obtained, and the product entity name vector and each three are calculated. The first word vector distance between the entity name vectors in the tuple vector, the second word vector distance between the product attribute vector and the entity relationship vector in each triple vector, according to the first word vector distance and The second word vector distance is to calculate the vector distance between the target vector and the vector of each triplet, and realize the determination of the vector distance between the target vector and the vector of each triplet.

在其中一个实施例中，目标向量与各三元组的向量之间的向量距离为第一词向量距离和第二词向量距离之和。In one of the embodiments, the vector distance between the target vector and the vector of each triple is the sum of the first word vector distance and the second word vector distance.

在其中一个实施例中，如图6所示，在S210之后，所述方法还包括：In one of the embodiments, as shown in FIG. 6, after S210, the method further includes:

S602：根据反馈的产品信息生成新的三元组，新的三元组中包括产品实体名称、产品属性以及产品信息；S602: Generate a new triplet according to the feedback product information, and the new triplet includes the product entity name, product attributes, and product information;

S604：根据新的三元组中的产品实体名称、产品属性以及产品信息，更新知识图谱库。S604: Update the knowledge graph database according to the product entity name, product attribute, and product information in the new triplet.

当知识图谱库中不存在目标三元组的向量时，根据反馈的产品信息生成新的三元组，三元组中包括产品实体名称、产品属性以及产品信息，根据新的三元组，更新知识图谱库。更新知识图谱库的方式包括：根据新的三元组中的产品实体名称查询知识图谱库，当知识图谱库中存在对应的产品实体名称时，生成对应的产品实体名称、产品信息以及产品属性的三元组。当知识图谱库中不存在对应的产品实体名称时，生成产品实体名称、产品属性以及产品信息的三元组。When the vector of the target triplet does not exist in the knowledge graph database, a new triplet is generated according to the feedback product information. The triplet includes the product entity name, product attributes and product information, and updates according to the new triplet Knowledge graph library. The way to update the knowledge graph database includes: query the knowledge graph database according to the product entity name in the new triplet, and when the corresponding product entity name exists in the knowledge graph database, generate the corresponding product entity name, product information and product attributes Triad. When the corresponding product entity name does not exist in the knowledge graph database, a triple of the product entity name, product attributes, and product information is generated.

上述实施例，根据反馈的产品信息生成新的三元组，新的三元组中包括产品实体名称、产品属性以及产品信息，根据新的三元组中的产品实体名称、产品属性以及产品信息，更新知识图谱库，实现了对知识图谱库的更新。In the above embodiment, a new triplet is generated based on the feedback product information. The new triplet includes the product entity name, product attribute, and product information. According to the product entity name, product attribute, and product information in the new triplet , Update the knowledge graph library, and realize the update of the knowledge graph library.

在其中一个实施例中，S604包括：In one of the embodiments, S604 includes:

S606：根据新的三元组中的产品实体名称查询知识图谱库；S606: Query the knowledge graph database according to the product entity name in the new triplet;

S608：当知识图谱库中存在对应的产品实体名称时，根据对应的产品实体名称、新的三元组中的产品属性以及产品信息，生成知识图谱库的三元组；S608: When the corresponding product entity name exists in the knowledge graph database, generate a triplet of the knowledge graph database according to the corresponding product entity name, product attributes in the new triplet, and product information;

S610：当知识图谱库中不存在对应的产品实体名称时，根据新的三元组中的产品实体名称、产品属性以及产品信息，生成知识图谱库的三元组。S610: When the corresponding product entity name does not exist in the knowledge graph database, generate a triple of the knowledge graph database according to the product entity name, product attribute, and product information in the new triple.

在其中一个实施例中，如图7所示，产品信息推送方法，还包括：In one of the embodiments, as shown in FIG. 7, the product information push method further includes:

S702：根据产品信息获取请求中携带的用户信息，获取用户历史请求数据和历史浏览数据；S702: Obtain user historical request data and historical browsing data according to the user information carried in the product information acquisition request;

S704：根据用户历史请求数据和历史浏览数据，确定用户偏好特征；S704: Determine user preference characteristics according to user historical request data and historical browsing data;

S706：根据用户偏好特征确定推荐产品信息；S706: Determine recommended product information according to user preference characteristics;

S708：推送推荐产品信息。S708: Push recommended product information.

服务器根据产品信息获取请求中携带的用户信息，获取用户历史请求数据和历史浏览数据，根据用户历史请求数据和历史浏览数据确定用户偏好特征，从而根据用户偏好特征确定推荐产品信息，推送推荐产品信息。The server obtains user historical request data and historical browsing data according to the user information carried in the product information obtaining request, determines user preference characteristics according to the user historical request data and historical browsing data, determines recommended product information according to user preference characteristics, and pushes recommended product information .

上述实施例，根据产品信息获取请求中携带的用户信息，获取用户历史请求数据和历史浏览数据，根据用户历史请求数据和历史浏览数据，确定推荐产品信息，推送推荐产品信息，实现了对推荐产品信息的推送。In the above embodiment, according to the user information carried in the product information acquisition request, the user history request data and the history browsing data are obtained, and the recommended product information is determined according to the user history request data and the history browsing data, and the recommended product information is pushed, so as to realize the recommended product Push of information.

应该理解的是，虽然图2-7的流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，这些步骤可以以其它的顺序执行。而且，图2-7中的至少一部分步骤可以包括多个子步骤或者多个阶段，这些子步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，这些子步骤或者阶段的执行顺序也不必然是依次进行，而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。It should be understood that although the various steps in the flowcharts of FIGS. 2-7 are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless specifically stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least part of the steps in Figures 2-7 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. These sub-steps or stages The execution order of is not necessarily performed sequentially, but may be performed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

在其中一个实施例中，如图8所示，提供了一种产品信息推送装置，包括：实体识别模块802、第一处理模块804、第二处理模块806、第一推送模块808和第二推送模块810，其中：In one of the embodiments, as shown in FIG. 8, a product information pushing device is provided, including: an entity recognition module 802, a first processing module 804, a second processing module 806, a first pushing module 808, and a second pushing Module 810, where:

实体识别模块802，用于接收产品信息获取请求，从产品信息获取请求中提取产品实体名称；The entity identification module 802 is used to receive the product information acquisition request, and extract the product entity name from the product information acquisition request;

第一处理模块804，用于根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据产品实体名称的词向量，生成产品信息获取请求的目标向量；The first processing module 804 is configured to determine the word vector of the product entity name according to the product entity name and the preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

第二处理模块806，用于计算目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，三元组包括两个实体名称以及两个实体名称之间的实体关系；The second processing module 806 is configured to calculate the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library, the triplet includes two entity names and the entity relationship between the two entity names;

第一推送模块808，用于当知识图谱库中存在目标三元组的向量时，将产品实体名称与目标三元组中的两个实体名称进行匹配，确定与产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将产品信息推送至产品信息获取请求的发送方，目标三元组的向量为与目标向量的向量距离在预设的向量距离阈值范围内的向量；及The first push module 808 is used to match the product entity name with two entity names in the target triple when there is a vector of the target triple in the knowledge graph database, and determine the entity name that matches the product entity name, According to the entity relationship of the entity name matching the product entity name, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is the vector distance between the target vector and the target vector. Set the vector within the distance threshold range; and

第二推送模块810，用于当知识图谱库中不存在目标三元组的向量时，根据产品信息获取请求生成服务提示，并将服务提示发送至客服终端，将客服终端反馈的产品信息推送至产品信息获取请求的发送方。The second push module 810 is used to generate a service reminder according to the product information acquisition request when there is no target triple vector in the knowledge graph database, and send the service reminder to the customer service terminal, and push the product information fed back by the customer service terminal to The sender of the product information acquisition request.

在其中一个实施例中，实体识别模块还用于接收产品信息获取请求，将产品信息获取请求转换为文字数据，对文字数据进行句法分析，确定文字数据的句法结构，将文字数据拆分为多个词语，确定拆分后的多个词语的词性，根据句法结构以及拆分后的多个词语的词性确定产品实体名称。In one of the embodiments, the entity recognition module is also used to receive the product information acquisition request, convert the product information acquisition request into text data, perform syntactic analysis on the text data, determine the syntactic structure of the text data, and split the text data into multiple Words, determine the part of speech of the split words, and determine the product entity name based on the syntactic structure and the part of speech of the split words.

在其中一个实施例中，实体识别模块还用于根据句法结构确定各句子中的主语、谓语以及宾语，根据句法结构以及拆分后的多个词语的词性筛选文字数据，筛选掉文字数据中的谓语、动词以及形容词，词性包括动词、形容词以及名词，根据筛选后的文字数据确定产品实体名称。In one of the embodiments, the entity recognition module is also used to determine the subject, predicate, and object in each sentence according to the syntactic structure, filter the text data according to the syntactic structure and the parts of speech of the split words, and filter out the text data Predicates, verbs, and adjectives. Parts of speech include verbs, adjectives, and nouns. The product entity name is determined based on the filtered text data.

在其中一个实施例中，第一处理模块还用于根据产品实体名称以及预设的词向量数据库，确定产品实体名称的词向量，根据预设的产品属性词库提取产品信息获取请求中的多个词语，根据预设的词向量数据库，确定提取出的各词语的词向量，根据提取出的各词语的词向量以及产品实体名称的词向量，生成产品信息获取请求的目标向量。In one of the embodiments, the first processing module is further configured to determine the word vector of the product entity name according to the product entity name and the preset word vector database, and extract the product information acquisition request according to the preset product attribute word database. According to the preset word vector database, determine the word vector of each word extracted, and generate the target vector of the product information acquisition request according to the word vector of each word extracted and the word vector of the product entity name.

在其中一个实施例中，第二处理模块还用于获取预设的知识图谱库中各三元组的向量中的实体名称向量以及实体关系向量，获取目标向量中的产品实体名称向量以及产品属性向量，计算产品实体名称向量与各三元组的向量中的实体名称向量之间的第一词向量距离，计算产品属性向量与各三元组的向量中的实体关系向量之间的第二词向量距离，根据第一词向量距离和第二词向量距离，计算目标向量与各三元组的向量之间的向量距离。In one of the embodiments, the second processing module is also used to obtain the entity name vector and entity relationship vector in the vector of each triple in the preset knowledge graph library, and obtain the product entity name vector and product attribute in the target vector Vector, calculate the first word vector distance between the product entity name vector and the entity name vector in the vector of each triple, and calculate the second word between the product attribute vector and the entity relationship vector in the vector of each triple The vector distance, according to the first word vector distance and the second word vector distance, calculate the vector distance between the target vector and the vector of each triplet.

在其中一个实施例中，产品信息推送装置还包括更新模块，更新模块用于根据反馈的产品信息生成新的三元组，新的三元组中包括产品实体名称、产品属性以及产品信息，根据新的三元组中的产品实体名称、产品属性以及产品信息，更新知识图谱库。In one of the embodiments, the product information pushing device further includes an update module, which is used to generate a new triplet according to the feedback product information. The new triplet includes the product entity name, product attributes, and product information, according to The product entity name, product attributes, and product information in the new triples are updated in the knowledge graph database.

在其中一个实施例中，更新模块还用于根据新的三元组中的产品实体名称查询知识图谱库，当知识图谱库中存在对应的产品实体名称时，根据对应的产品实体名称、新的三元组中的产品属性以及产品信息，生成知识图谱库的三元组，当知识图谱库中不存在对应的产品实体名称时，根据新的三元组中的产品实体名称、产品属性以及产品信息，生成知识图谱库的三元组。In one of the embodiments, the update module is also used to query the knowledge graph database according to the product entity name in the new triplet, and when the corresponding product entity name exists in the knowledge graph database, according to the corresponding product entity name, new The product attribute and product information in the triplet are generated to generate the triplet of the knowledge graph database. When the corresponding product entity name does not exist in the knowledge graph database, the product entity name, product attribute and product in the new triplet Information, generating triples of the knowledge graph library.

在其中一个实施例中，产品信息推送装置还包括推荐模块，推荐模块用于根据产品信息获取请求中携带的用户信息，获取用户历史请求数据和历史浏览数据，根据用户历史请求数据和历史浏览数据，确定用户偏好特征，根据用户偏好特征确定推荐产品信息，推送推荐产品信息。In one of the embodiments, the product information pushing device further includes a recommendation module. The recommendation module is used to obtain user information carried in the request according to product information, obtain user history request data and historical browsing data, and obtain user history request data and historical browsing data according to user history request data and historical browsing data. , Determine user preference characteristics, determine recommended product information according to user preference characteristics, and push recommended product information.

关于产品信息推送装置的具体限定可以参见上文中对于产品信息推送方法的限定，在此不再赘述。上述产品信息推送装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。For the specific limitation of the product information pushing device, please refer to the above limitation on the product information pushing method, which is not repeated here. Each module in the above product information pushing device can be implemented in whole or in part by software, hardware, and a combination thereof. The foregoing modules may be embedded in the form of hardware or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the foregoing modules.

在一个实施例中，提供了一种计算机设备，该计算机设备可以是服务器，其内部结构图可以如图9所示。该计算机设备包括通过***总线连接的处理器、存储器、网络接口和数据库。其中，该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作***、计算机可读指令和数据库。该内存储器为非易失性存储介质中的操作***和计算机可读指令的运行提供环境。该计算机设备的数据库用于存储知识图谱数据以及词向量数据库。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种产品信息推送方法。In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 9. The computer equipment includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer readable instructions, and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile storage medium. The database of the computer equipment is used to store the knowledge map data and the word vector database. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer-readable instruction is executed by the processor to realize a product information push method.

本领域技术人员可以理解，图9中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。Those skilled in the art can understand that the structure shown in FIG. 9 is only a block diagram of part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.

一种计算机设备，包括存储器和一个或多个处理器，存储器中储存有计算机可读指令，计算机可读指令被处理器执行时，使得一个或多个处理器执行以下步骤：A computer device includes a memory and one or more processors. The memory stores computer readable instructions. When the computer readable instructions are executed by the processor, the one or more processors execute the following steps:

本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机可读指令来指令相关的硬件来完成，所述的计算机可读指令可存储于一非易失性计算机可读取存储介质中，该计算机可读指令在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限，RAM以多种形式可得，诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接 RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through computer-readable instructions, which can be stored in a non-volatile computer. In a readable storage medium, when the computer-readable instructions are executed, they may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Channel (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction between the combinations of these technical features, they should It is considered as the range described in this specification.

以上所述实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请专利的保护范围应以所附权利要求为准。The above-mentioned embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can be made, and these all fall within the protection scope of this application. Therefore, the scope of protection of the patent of this application shall be subject to the appended claims.

Claims

一种产品信息推送方法，包括：A product information push method, including:

接收产品信息获取请求，从所述产品信息获取请求中提取产品实体名称；Receive a product information acquisition request, and extract the product entity name from the product information acquisition request;

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量，根据所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

计算所述目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，所述三元组包括两个实体名称以及两个实体名称之间的实体关系；Calculating the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library, the triplet including two entity names and the entity relationship between the two entity names;

当所述知识图谱库中存在目标三元组的向量时，将所述产品实体名称与所述目标三元组中的两个实体名称进行匹配，确定与所述产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将所述产品信息推送至所述产品信息获取请求的发送方，所述目标三元组的向量为与所述目标向量的向量距离在预设的向量距离阈值范围内的向量；及When the vector of target triples exists in the knowledge graph library, the product entity name is matched with two entity names in the target triple, and the entity name that matches the product entity name is determined, And according to the entity relationship of the entity name matching the product entity name, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is The vector whose vector distance of the target vector is within the preset vector distance threshold range; and

当所述知识图谱库中不存在所述目标三元组的向量时，根据所述产品信息获取请求生成服务提示，并将所述服务提示发送至客服终端，将所述客服终端反馈的产品信息推送至所述产品信息获取请求的发送方。When the vector of the target triplet does not exist in the knowledge graph library, a service prompt is generated according to the product information acquisition request, and the service prompt is sent to the customer service terminal, and the product information fed back by the customer service terminal Push to the sender of the product information acquisition request.
根据权利要求1所述的方法，其特征在于，所述接收产品信息获取请求，从所述产品信息获取请求中提取产品实体名称，包括：The method according to claim 1, wherein the receiving the product information acquisition request and extracting the product entity name from the product information acquisition request comprises:

接收产品信息获取请求，将所述产品信息获取请求转换为文字数据；Receiving a product information acquisition request, and converting the product information acquisition request into text data;

对所述文字数据进行句法分析，确定文字数据的句法结构；Perform syntactic analysis on the text data to determine the syntactic structure of the text data;

将文字数据拆分为多个词语，确定拆分后的多个词语的词性；及Split the text data into multiple words and determine the part of speech of the multiple words after the split; and

根据所述句法结构以及拆分后的多个词语的词性确定产品实体名称。The product entity name is determined according to the syntactic structure and the part of speech of multiple words after splitting.
根据权利要求2所述的方法，其特征在于，所述根据所述句法结构以及拆分后的多个词语的词性确定产品实体名称，包括：The method according to claim 2, wherein the determining the product entity name according to the syntactic structure and the parts of speech of the split words comprises:

根据所述句法结构确定各句子中的主语、谓语以及宾语；Determine the subject, predicate and object in each sentence according to the syntactic structure;

根据所述句法结构以及拆分后的多个词语的词性筛选所述文字数据，筛选掉所述文字数据中的谓语、动词以及形容词，所述词性包括动词、形容词以及名词；及Filter the text data according to the syntactic structure and the parts of speech of the split words, and filter out predicates, verbs and adjectives in the text data, and the parts of speech include verbs, adjectives and nouns; and

根据筛选后的文字数据确定产品实体名称。Determine the product entity name based on the filtered text data.
根据权利要求1所述的方法，其特征在于，所述根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量，根据所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量，包括：The method according to claim 1, wherein the word vector of the product entity name is determined according to the product entity name and a preset word vector database, and the word vector of the product entity name is generated according to the word vector of the product entity name The target vector of the product information acquisition request includes:

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database;

根据预设的产品属性词库提取所述产品信息获取请求中的多个词语，根据所述预设的词向量数据库，确定提取出的各词语的词向量；及Extract multiple words in the product information acquisition request according to a preset product attribute word database, and determine the word vector of each word extracted according to the preset word vector database; and

根据所述提取出的各词语的词向量以及所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量。According to the word vector of each word extracted and the word vector of the product entity name, the target vector of the product information acquisition request is generated.
根据权利要求1所述的方法，其特征在于，所述计算所述目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，包括：The method according to claim 1, wherein the calculating the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library comprises:

获取预设的知识图谱库中各三元组的向量中的实体名称向量以及实体关系向量；Obtain the entity name vector and entity relationship vector in the vector of each triplet in the preset knowledge graph library;

获取所述目标向量中的产品实体名称向量以及产品属性向量；Acquiring a product entity name vector and a product attribute vector in the target vector;

计算所述产品实体名称向量与各所述三元组的向量中的实体名称向量之间的第一词向量距离；Calculating the first word vector distance between the product entity name vector and the entity name vector in each of the triplet vectors;

计算所述产品属性向量与各所述三元组的向量中的实体关系向量之间的第二词向量距离；及Calculating the second word vector distance between the product attribute vector and the entity relationship vector in the vector of each of the triples; and

根据所述第一词向量距离和所述第二词向量距离，计算所述目标向量与各所述三元组的向量之间的向量距离。According to the first word vector distance and the second word vector distance, the vector distance between the target vector and the vector of each of the triples is calculated.
根据权利要求5所述的方法，其特征在于，所述目标向量与各所述三元组的向量之间的向量距离为所述第一词向量距离和所述第二词向量距离之和。The method according to claim 5, wherein the vector distance between the target vector and the vector of each of the triples is the sum of the first word vector distance and the second word vector distance.
根据权利要求1所述的方法，其特征在于，在所述将所述客服终端反馈的产品信息推送至所述产品信息获取请求的发送方之后，所述方法还包括：The method according to claim 1, wherein after the product information fed back by the customer service terminal is pushed to the sender of the product information acquisition request, the method further comprises:

根据所述反馈的产品信息生成新的三元组，所述新的三元组中包括产品实体名称、产品属性以及产品信息；及Generate a new triplet according to the feedback product information, the new triplet including the product entity name, product attributes, and product information; and

根据新的三元组中的产品实体名称、产品属性以及产品信息，更新所述知识图谱库。The knowledge graph database is updated according to the product entity name, product attribute, and product information in the new triplet.
根据权利要求7所述的方法，其特征在于，所述更新所述知识图谱库，包括：The method according to claim 7, wherein said updating said knowledge graph library comprises:

根据所述新的三元组中的产品实体名称查询所述知识图谱库；Query the knowledge graph database according to the product entity name in the new triplet;

当所述知识图谱库中存在对应的产品实体名称时，根据所述对应的产品实体名称、所述新的三元组中的产品属性以及产品信息，生成所述知识图谱库的三元组；及When the corresponding product entity name exists in the knowledge graph database, generate the triplet of the knowledge graph database according to the corresponding product entity name, product attributes in the new triplet, and product information; and

当所述知识图谱库中不存在对应的产品实体名称时，根据所述新的三元组中的产品实体名称、产品属性以及产品信息，生成所述知识图谱库的三元组。When the corresponding product entity name does not exist in the knowledge graph database, a triple of the knowledge graph database is generated according to the product entity name, product attribute, and product information in the new triple.
根据权利要求1所述的方法，其特征在于，还包括：The method according to claim 1, further comprising:

根据产品信息获取请求中携带的用户信息，获取用户历史请求数据和历史浏览数据；Obtain user historical request data and historical browsing data according to the user information carried in the product information acquisition request;

根据所述用户历史请求数据和所述历史浏览数据，确定用户偏好特征；Determining user preference characteristics according to the user historical request data and the historical browsing data;

根据所述用户偏好特征确定推荐产品信息；及Determine recommended product information according to the user preference characteristics; and

推送所述推荐产品信息。Push the recommended product information.
一种产品信息推送装置，包括：A product information push device, including:

实体识别模块，用于接收产品信息获取请求，从所述产品信息获取请求中提取产品实体名称；The entity identification module is used to receive the product information acquisition request, and extract the product entity name from the product information acquisition request;

第一处理模块，用于根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量，根据所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量；The first processing module is configured to determine the word vector of the product entity name according to the product entity name and a preset word vector database, and generate the target of the product information acquisition request according to the word vector of the product entity name vector;

第二处理模块，用于计算所述目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，所述三元组包括两个实体名称以及两个实体名称之间的实体关系；The second processing module is used to calculate the vector distance between the target vector and the vector of each triplet in the preset knowledge graph database. The triplet includes two entity names and the distance between the two entity names. Entity relationship

第一推送模块，用于当所述知识图谱库中存在目标三元组的向量时，将所述产品实体名称与所述目标三元组中的两个实体名称进行匹配，确定与所述产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将所述产品信息推送至所述产品信息获取请求的发送方，所述目标三元组的向量为与所述目标向量的向量距离在预设的向量距离阈值范围内的向量；及The first push module is used to match the product entity name with the two entity names in the target triple when there is a vector of the target triple in the knowledge graph library, and determine that it matches the product The entity name matches the entity name, and the corresponding product information is determined according to the entity relationship of the entity name that matches the product entity name, and the product information is pushed to the sender of the product information acquisition request, the target three The vector of the tuple is a vector whose vector distance to the target vector is within a preset vector distance threshold range; and

第二推送模块，用于当所述知识图谱库中不存在所述目标三元组的向量时，根据所述产品信息获取请求生成服务提示，并将所述服务提示发送至客服终端，将所述客服终端反馈的产品信息推送至所述产品信息获取请求的发送方。The second push module is configured to generate a service prompt according to the product information acquisition request when the target triple vector does not exist in the knowledge graph database, and send the service prompt to the customer service terminal, and send all The product information fed back by the customer service terminal is pushed to the sender of the product information acquisition request.
根据权利要求10所述的装置，其特征在于，所述实体识别模块还用于接收产品信息获取请求，将所述产品信息获取请求转换为文字数据，对所述文字数据进行句法分析，确定文字数据的句法结构，将文字数据拆分为多个词语，确定拆分后的多个词语的词性，根据所述句法结构以及拆分后的多个词语的词性确定产品实体名称。The device according to claim 10, wherein the entity recognition module is further configured to receive a product information acquisition request, convert the product information acquisition request into text data, perform syntactic analysis on the text data, and determine the text The syntactic structure of the data splits the text data into multiple words, determines the part of speech of the split words, and determines the product entity name according to the syntactic structure and the part of speech of the split words.
根据权利要求10所述的装置，其特征在于，所述实体识别模块还用于根据所述句法结构确定各句子中的主语、谓语以及宾语，根据所述句法结构以及拆分后的多个词语的词性筛选所述文字数据，筛选掉所述文字数据中的谓语、动词以及形容词，所述词性包括动词、形容词以及名词，根据筛选后的文字数据确定产品实体名称。The device according to claim 10, wherein the entity recognition module is further configured to determine the subject, the predicate, and the object in each sentence according to the syntactic structure, and according to the syntactic structure and the split words The part of speech filters the text data to filter out predicates, verbs, and adjectives in the text data. The part of speech includes verbs, adjectives, and nouns, and the product entity name is determined according to the filtered text data.
一种计算机设备，包括存储器及一个或多个处理器，所述存储器中储存有计算机可读指令，所述计算机可读指令被所述一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：A computer device includes a memory and one or more processors. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the one or more processors, the one or more Each processor performs the following steps:

接收产品信息获取请求，从所述产品信息获取请求中提取产品实体名称；Receive a product information acquisition request, and extract the product entity name from the product information acquisition request;

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量，根据所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

计算所述目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，所述三元组包括两个实体名称以及两个实体名称之间的实体关系；Calculating the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library, the triplet including two entity names and the entity relationship between the two entity names;

当所述知识图谱库中存在目标三元组的向量时，将所述产品实体名称与所述目标三元组中的两个实体名称进行匹配，确定与所述产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将所述产品信息推送至所述产品信息获取请求的发送方，所述目标三元组的向量为与所述目标向量的向量距离在预设的向量距离阈值范围内的向量；及When the vector of target triples exists in the knowledge graph library, the product entity name is matched with two entity names in the target triple, and the entity name that matches the product entity name is determined, And according to the entity relationship of the entity name matching the product entity name, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is The vector whose vector distance of the target vector is within the preset vector distance threshold range; and

当所述知识图谱库中不存在所述目标三元组的向量时，根据所述产品信息获取请求生成服务提示，并将所述服务提示发送至客服终端，将所述客服终端反馈的产品信息推送至所述产品信息获取请求的发送方。When the vector of the target triplet does not exist in the knowledge graph library, a service prompt is generated according to the product information acquisition request, and the service prompt is sent to the customer service terminal, and the product information fed back by the customer service terminal Push to the sender of the product information acquisition request.
根据权利要求13所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时还执行以下步骤：The computer device according to claim 13, wherein the processor further executes the following steps when executing the computer-readable instruction:

接收产品信息获取请求，将所述产品信息获取请求转换为文字数据；Receiving a product information acquisition request, and converting the product information acquisition request into text data;

对所述文字数据进行句法分析，确定文字数据的句法结构；Perform syntactic analysis on the text data to determine the syntactic structure of the text data;

将文字数据拆分为多个词语，确定拆分后的多个词语的词性；及Split the text data into multiple words and determine the part of speech of the multiple words after the split; and

根据所述句法结构以及拆分后的多个词语的词性确定产品实体名称。The product entity name is determined according to the syntactic structure and the part of speech of multiple words after splitting.
根据权利要求13所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时还执行以下步骤：The computer device according to claim 13, wherein the processor further executes the following steps when executing the computer-readable instruction:

根据所述句法结构确定各句子中的主语、谓语以及宾语；Determine the subject, predicate and object in each sentence according to the syntactic structure;

根据所述句法结构以及拆分后的多个词语的词性筛选所述文字数据，筛选掉所述文字数据中的谓语、动词以及形容词，所述词性包括动词、形容词以及名词；及Filter the text data according to the syntactic structure and the parts of speech of the split words, and filter out predicates, verbs and adjectives in the text data, and the parts of speech include verbs, adjectives and nouns; and

根据筛选后的文字数据确定产品实体名称。Determine the product entity name based on the filtered text data.
根据权利要求13所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时还执行以下步骤：The computer device according to claim 13, wherein the processor further executes the following steps when executing the computer-readable instruction:

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database;

根据预设的产品属性词库提取所述产品信息获取请求中的多个词语，根据所述预设的词向量数据库，确定提取出的各词语的词向量；及Extract multiple words in the product information acquisition request according to a preset product attribute word database, and determine the word vector of each word extracted according to the preset word vector database; and

根据所述提取出的各词语的词向量以及所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量。According to the word vector of each word extracted and the word vector of the product entity name, the target vector of the product information acquisition request is generated.
一个或多个存储有计算机可读指令的非易失性计算机可读存储介质，所述计算机可读指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：One or more non-volatile computer-readable storage media storing computer-readable instructions, which when executed by one or more processors, cause the one or more processors to perform the following steps:

接收产品信息获取请求，从所述产品信息获取请求中提取产品实体名称；Receive a product information acquisition request, and extract the product entity name from the product information acquisition request;

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量，根据所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database, and generate the target vector of the product information acquisition request according to the word vector of the product entity name;

计算所述目标向量与预设的知识图谱库中各三元组的向量之间的向量距离，所述三元组包括两个实体名称以及两个实体名称之间的实体关系；Calculating the vector distance between the target vector and the vector of each triplet in the preset knowledge graph library, the triplet including two entity names and the entity relationship between the two entity names;

当所述知识图谱库中存在目标三元组的向量时，将所述产品实体名称与所述目标三元组中的两个实体名称进行匹配，确定与所述产品实体名称匹配的实体名称，并根据与产品实体名称匹配的实体名称所在的实体关系，确定对应的产品信息，并将所述产品信息推送至所述产品信息获取请求的发送方，所述目标三元组的向量为与所述目标向量的向量距离在预设的向量距离阈值范围内的向量；及When the vector of target triples exists in the knowledge graph library, the product entity name is matched with two entity names in the target triple, and the entity name that matches the product entity name is determined, And according to the entity relationship of the entity name matching the product entity name, the corresponding product information is determined, and the product information is pushed to the sender of the product information acquisition request. The vector of the target triplet is The vector whose vector distance of the target vector is within the preset vector distance threshold range; and

当所述知识图谱库中不存在所述目标三元组的向量时，根据所述产品信息获取请求生成服务提示，并将所述服务提示发送至客服终端，将所述客服终端反馈的产品信息推送至所述产品信息获取请求的发送方。When the vector of the target triplet does not exist in the knowledge graph library, a service prompt is generated according to the product information acquisition request, and the service prompt is sent to the customer service terminal, and the product information fed back by the customer service terminal Push to the sender of the product information acquisition request.
根据权利要求17所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：18. The storage medium of claim 17, wherein the following steps are further executed when the computer-readable instructions are executed by the processor:

接收产品信息获取请求，将所述产品信息获取请求转换为文字数据；Receiving a product information acquisition request, and converting the product information acquisition request into text data;

对所述文字数据进行句法分析，确定文字数据的句法结构；Perform syntactic analysis on the text data to determine the syntactic structure of the text data;

将文字数据拆分为多个词语，确定拆分后的多个词语的词性；及Split the text data into multiple words and determine the part of speech of the multiple words after the split; and

根据所述句法结构以及拆分后的多个词语的词性确定产品实体名称。The product entity name is determined according to the syntactic structure and the part of speech of multiple words after splitting.
根据权利要求17所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：18. The storage medium of claim 17, wherein the following steps are further executed when the computer-readable instructions are executed by the processor:

根据所述句法结构确定各句子中的主语、谓语以及宾语；Determine the subject, predicate and object in each sentence according to the syntactic structure;

根据所述句法结构以及拆分后的多个词语的词性筛选所述文字数据，筛选掉所述文字数据中的谓语、动词以及形容词，所述词性包括动词、形容词以及名词；及Filter the text data according to the syntactic structure and the parts of speech of the split words, and filter out predicates, verbs and adjectives in the text data, and the parts of speech include verbs, adjectives and nouns; and

根据筛选后的文字数据确定产品实体名称。Determine the product entity name based on the filtered text data.
根据权利要求17所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：18. The storage medium of claim 17, wherein the following steps are further executed when the computer-readable instructions are executed by the processor:

根据所述产品实体名称以及预设的词向量数据库，确定所述产品实体名称的词向量；Determine the word vector of the product entity name according to the product entity name and a preset word vector database;

根据预设的产品属性词库提取所述产品信息获取请求中的多个词语，根据所述预设的词向量数据库，确定提取出的各词语的词向量；及Extract multiple words in the product information acquisition request according to a preset product attribute word database, and determine the word vector of each word extracted according to the preset word vector database; and

根据所述提取出的各词语的词向量以及所述产品实体名称的词向量，生成所述产品信息获取请求的目标向量。According to the word vector of each word extracted and the word vector of the product entity name, the target vector of the product information acquisition request is generated.