Insurance proposal generation method and device, computer equipment and storage medium

Abstract

The embodiment of the invention provides an insurance proposal generation method and device, computer equipment and a storage medium, relates to the technical field of personalized data generation, and is suitable for the field of financial science and technology. The method comprises the steps of obtaining basic information of a user, insurance question and answer data and current behavior data of the user; predicting insurance preference prediction data of the target user according to the insurance question and answer data and the current behavior data of the user; acquiring a historical insurance proposal, current insurance product information and insurance claim settlement data; extracting candidate insurance key fields from the historical insurance proposal, the current insurance product information and the insurance claim settlement data according to a key field extraction rule; screening out a target insurance key field from the candidate insurance key fields according to the user basic information and the insurance preference prediction data; and performing proposal generation through the user basic information and the target insurance key field to obtain a target insurance proposal. According to the embodiment of the invention, the professional requirement and manpower during the generation of the insurance proposal can be reduced.

Description

Technical Field

[0001] The present application relates to the field of personalized data generation technology, which is applicable to the field of financial technology, and in particular to a method and device for generating an insurance proposal, a computer device, and a storage medium. Background Art

[0002] In the insurance industry, insurance proposals need to be generated based on the customer's current situation, needs, and product conditions in the insurance market, and can provide customers with professional insurance advice. In related technologies, insurance proposals are mainly generated manually by insurance agents after talking with customers, based on customer needs, current situation, and insurance product-related knowledge. This not only requires insurance agents to have rich experience and knowledge reserves in insurance, but also requires insurance agents to spend a lot of time and energy. Therefore, how to reduce the professional requirements and manpower for insurance agents to prepare insurance proposals has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of the embodiments of the present application is to propose an insurance proposal generation method and apparatus, a computer device and a storage medium, aiming to realize the automatic generation of insurance proposals and reduce the professional requirements and manpower when generating insurance proposals.

[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a method for generating an insurance proposal, the method comprising:

[0005] Obtaining basic user information, insurance question and answer data, and current user behavior data of the target user; wherein the current user behavior data represents the operation behavior of the target user on the insurance business platform;

[0006] Predicting the preferences of the target user based on the insurance question-and-answer data and the user's current behavior data to obtain insurance preference prediction data;

[0007] Access historical insurance proposals, current insurance product information, and insurance claims data;

[0008] Extract key fields from the historical insurance proposal, the current insurance product information and the insurance claim data according to preset key field extraction rules to obtain candidate insurance key fields; wherein the candidate insurance key fields are key contents for constructing the insurance proposal;

[0009] Filtering target insurance key fields from the candidate insurance key fields according to the user basic information and the insurance preference prediction data;

[0010] The target insurance proposal is obtained by generating a proposal based on a preset original proposal generation model, the user basic information and the target insurance key fields.

[0011] In some embodiments, the key field extraction rules include: field splitting rules and key field classification rules;

[0012] The key field extraction rules are used to extract the key fields of the historical insurance proposal, the current insurance product information and the insurance claim data to obtain candidate insurance key fields, including:

[0013] Perform field splitting on the historical insurance proposal, the current insurance product information and the insurance claim data according to the field splitting rule to obtain candidate fields;

[0014] Performing category identification on each of the candidate fields to obtain a field category;

[0015] The candidate fields are constructed into key fields according to the key field classification rules and the field categories to obtain the candidate insurance key fields.

[0016] In some embodiments, the step of selecting a target insurance key field from the candidate insurance key fields according to the user basic information and the insurance preference prediction data includes:

[0017] Acquire field tag information of the candidate insurance key field; wherein the field tag information includes: candidate user information and candidate insurance attribute information;

[0018] The candidate insurance key fields are screened according to the user basic information, the insurance preference prediction data, the candidate user information and the candidate insurance attribute information to obtain the target insurance key fields.

[0019] In some embodiments, the screening of the candidate insurance key fields according to the user basic information, the insurance preference prediction data, the candidate user information and the candidate insurance attribute information to obtain the target insurance key fields includes:

[0020] Performing vector conversion on the user basic information and the insurance preference prediction data to obtain a query vector;

[0021] Performing vector conversion on the candidate user information and the candidate insurance attribute information to obtain a reference vector;

[0022] Calculating similarity between the query vector and the reference vector to obtain vector similarity;

[0023] The candidate insurance key fields are screened according to the vector similarity to obtain the target insurance key fields.

[0024] In some embodiments, the step of performing preference prediction on the target user based on the insurance question-and-answer data and the user's current behavior data to obtain insurance preference prediction data includes:

[0025] Extracting features from the insurance question and answer data to obtain insurance question and answer features;

[0026] Extracting features from the user's current behavior data to obtain features of the user's current behavior;

[0027] Preference prediction is performed through a preset preference prediction model, the insurance question and answer characteristics, and the user's current behavior characteristics to obtain the insurance preference prediction data.

[0028] In some embodiments, after generating a proposal by using a preset original proposal generation model, the user basic information and the target insurance key fields to obtain a target insurance proposal, the method further includes:

[0029] Performing a quality assessment on the target insurance proposal according to a preset verification insurance proposal and a preset quality assessment rule to obtain quality assessment data; wherein the quality assessment data represents the degree of matching between the target insurance proposal and the verification insurance proposal, and the degree of matching between the target insurance proposal and the quality assessment rule;

[0030] Optimizing the original proposal generation model according to the quality assessment data to obtain a target proposal generation model;

[0031] The target insurance proposal is optimized through the target proposal generation model, the user basic information and the target insurance key fields to obtain an updated insurance proposal.

[0032] In some embodiments, the step of performing a quality assessment on the target insurance proposal according to a preset verification insurance proposal and a preset quality assessment rule to obtain quality assessment data includes:

[0033] Calculate the matching degree based on the verification insurance proposal and the target insurance proposal to obtain the content matching degree;

[0034] Extracting evaluation reference indicator data from the target insurance proposal; wherein the evaluation reference indicator data includes: insurance product information, insurance service information and insured user information;

[0035] Calculating the matching degree of the insurance product information, the insurance service information and the insured user information according to the quality assessment rule to obtain the information matching degree;

[0036] The content matching degree and the information matching degree are concatenated to obtain the quality assessment data.

[0037] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application provides an insurance proposal generation device, the device comprising:

[0038] A user data acquisition module is used to acquire basic user information, insurance question and answer data, and user current behavior data of a target user; wherein the user current behavior data represents the operation behavior of the target user on the insurance business platform;

[0039] A preference prediction module, used to predict the preference of the target user based on the insurance question-and-answer data and the user's current behavior data, to obtain insurance preference prediction data;

[0040] The historical data acquisition module is used to obtain historical insurance proposals, current insurance product information and insurance claims data;

[0041] A key field splitting module is used to split the historical insurance proposal, the current insurance product information and the insurance claim data into key fields according to a preset key field extraction rule to obtain candidate insurance key fields;

[0042] A key field screening module, used for screening target insurance key fields from the candidate insurance key fields according to the user basic information and the insurance preference prediction data;

[0043] The proposal screening module is used to generate a proposal through a preset original proposal generation model, the user basic information and the target insurance key fields to obtain a target insurance proposal.

[0044] To achieve the above-mentioned purpose, the third aspect of an embodiment of the present application proposes a computer device, which includes a memory and a processor, the memory stores a computer program, and the processor implements the method described in the first aspect when executing the computer program.

[0045] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0046] The insurance proposal generation method and device, computer equipment and storage medium proposed in the present application predict the insurance preference prediction data of the target object based on the current user behavior data and insurance question and answer data of the target user on the insurance business platform, and then split the historical insurance proposals, current insurance product information and insurance claims data on the insurance business platform into candidate insurance key fields for constructing insurance proposals in advance. Therefore, when constructing an insurance proposal, the target insurance key fields that can be used to construct the insurance proposal are screened from the candidate insurance key fields based on the user's basic information and insurance preference prediction data, and the target insurance key fields are combined into the target insurance proposal using the original proposal generation model and the user's basic information. Therefore, by automatically generating insurance proposals, and the generation process will be constructed based on the basic information and insurance preferences of the target user, not only the professional requirements and manpower for the generation of insurance proposals are reduced, but also insurance proposals that meet the target user can be generated, thereby improving the quality of insurance services. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of a method for generating an insurance proposal provided in an embodiment of the present application;

[0048] Figure 2 yes Figure 1 Flow chart of step S102 in FIG.

[0049] Figure 3 yes Figure 1 Flow chart of step S104 in FIG.

[0050] Figure 4 yes Figure 1 Flow chart of step S105 in FIG.

[0051] Figure 5 yes Figure 4 Flow chart of step S402 in FIG.

[0052] Figure 6 is a flow chart of a method for generating an insurance proposal provided by another embodiment of the present application;

[0053] Figure 7 yes Figure 6 Flow chart of step S601 in FIG.

[0054] Figure 8 It is a schematic diagram of an interface for generating a target insurance proposal in the insurance proposal generating method provided in an embodiment of the present application;

[0055] Fig. 9 It is a model structure diagram of the insurance proposal generation method provided in the embodiment of the present application;

[0056] Fig.10It is a structural schematic diagram of an insurance proposal generating device provided in an embodiment of the present application;

[0057] Fig.11 It is a schematic diagram of the hardware structure of the computer device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. 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.

[0059] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar users and are not necessarily used to describe a specific order or sequence.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0061] First, some nouns involved in this application are analyzed:

[0062] Artificial intelligence (AI) is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. AI is a branch of computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can respond in a similar way to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing and expert systems. AI can simulate the information process of human consciousness and thinking. AI is also a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0063] Large Language Model (LLM): is an AI model designed to understand and generate human language. Large language models are trained on large amounts of text data and can perform a wide range of tasks, including text summarization, translation, sentiment analysis, etc. LLMs are characterized by their large size and contain billions of parameters, which help large language models learn complex patterns in language data. These models are usually based on deep learning architectures such as transformers, which helps large language models achieve impressive performance on various natural language processing (NLP) tasks.

[0064] prompt: In the field of artificial intelligence (AI), "prompt" refers to text or instructions that provide input to a model to guide it to produce a specific output. It is a text paragraph provided by the user when interacting with the model, which is used to describe the information, answer, text, etc. that the user wants to obtain from the model. The purpose of the prompt is to guide the model to produce the desired response so that the generated output can be better controlled.

[0065] Insurance proposal: also known as insurance plan, refers to a written material in which insurance practitioners recommend suitable insurance products to customers and design the best insurance plan based on the customers' own financial status and financial management requirements, in order to seek the maximum insurance benefits for customers, while also helping customers understand and accept insurance products.

[0066] With the development of science and technology, the traditional business model of the insurance industry faces problems such as low efficiency and information asymmetry. For example, in the scenario of insurance proposal generation, insurance proposal generation is mainly done by insurance agents who talk with customers and prepare insurance proposals based on customer needs to introduce insurance products to customers through insurance proposals. Therefore, the generation of insurance proposals by insurance agents not only requires insurance agents to have rich experience and knowledge reserves, but also requires insurance agents to spend a lot of time and energy.

[0067] Based on this, the embodiment of the present application provides an insurance proposal generation method and device, a computer device and a storage medium, which predicts the insurance preference prediction data of the target object based on the current user behavior data and insurance question and answer data of the target user on the insurance business platform, and then splits the historical insurance proposals, current insurance product information and insurance claims data on the insurance business platform into candidate insurance key fields for constructing insurance proposals in advance. Therefore, when constructing an insurance proposal, the target insurance key fields that can construct the insurance proposal are screened from the candidate insurance key fields according to the user's basic information and insurance preference prediction data, and the original proposal generation model and the user's basic information are used to combine the target insurance key fields into a target insurance proposal. Therefore, by automatically generating insurance proposals, and the generation process will be constructed based on the basic information and insurance preferences of the target user, not only the professional requirements and manpower for the generation of insurance proposals are reduced, but also insurance proposals that meet the target user can be generated, thereby improving the quality of insurance services.

[0068] The insurance proposal generation method and apparatus, computer equipment and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the insurance proposal generation method in the embodiments of the present application is described.

[0069] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0070] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0071] The insurance proposal generation method provided in the embodiment of the present application relates to the field of personalized data generation technology and is applicable to the field of financial technology. The insurance proposal generation method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the insurance proposal generation method, etc., but is not limited to the above forms.

[0072] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer computer devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, users, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0073] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to the user's identity or characteristics such as user information, user current behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0074] Figure 1 is an optional flowchart of the method for generating an insurance proposal provided in an embodiment of the present application. Figure 1The method may include but is not limited to steps S101 to S106.

[0075] Step S101, obtaining basic user information, insurance question and answer data, and current user behavior data of a target user; wherein the current user behavior data represents the operation behavior of the target user on the insurance business platform;

[0076] Step S102, predicting the preferences of target users based on the insurance question-and-answer data and the user's current behavior data, to obtain insurance preference prediction data;

[0077] Step S103, obtaining historical insurance proposals, current insurance product information and insurance claims data;

[0078] Step S104, splitting the historical insurance proposals, current insurance product information and insurance claim data into key fields according to preset key field extraction rules to obtain candidate insurance key fields;

[0079] Step S105, selecting target insurance key fields from candidate insurance key fields according to user basic information and insurance preference prediction data;

[0080] Step S106, generating a proposal through the preset original proposal generation model, user basic information and target insurance key fields to obtain a target insurance proposal.

[0081] Steps S101 to S106 shown in the embodiment of the present application predict the insurance preference of the target user based on the current user behavior data and insurance question and answer data of the target user on the insurance business platform to obtain insurance preference prediction data. Before generating an insurance proposal, candidate insurance key fields for constructing an insurance proposal are extracted in advance from existing historical insurance proposals, current insurance product information, and insurance claims data. Therefore, in the process of automatically generating an insurance proposal, target insurance key fields that meet the insurance preferences and basic user information of the target user are extracted from the candidate insurance key fields, and the target insurance key fields are combined into a target insurance proposal using the original proposal generation model and the basic user information. Therefore, by automatically generating an insurance proposal using a model, and the insurance proposal generation process is completed in combination with the target user's insurance preferences, personal information, and historical insurance-related data, not only the professional requirements and manpower for generating an insurance proposal are reduced, but also an insurance proposal that meets the target user is constructed to improve the quality of insurance services.

[0082] In step S101 of some embodiments, the basic user information is the basic attribute information of the target user, specifically including: the target user's age, gender, income, occupation, risk tolerance and disease information, etc., and this embodiment does not impose specific restrictions on the content of the basic user information. Insurance Q&A data is obtained by using an intelligent insurance Q&A robot to conduct Q&A with the target user before the insurance proposal is generated, and the data generated during the Q&A period is collected as insurance Q&A data. The user's current behavior data is the behavior data of the target user on the insurance business platform, which is formed by browsing, clicking and purchasing on the insurance business platform, and the user's current behavior data includes the user's current browsing behavior data, policy duration data, etc. This embodiment does not impose specific restrictions on the content of the user's current behavior data.

[0083] Specifically, the target user's basic information, insurance question and answer data, and user's current behavior data can also be generated through face-to-face or online communication between the insurance agent and the target user, and the audio and video data can be analyzed and learned. In addition, a virtual insurance agent robot can also be trained through the audio and video data provided by professional insurance agents during customer meetings, and the virtual insurance agent robot can replace professional insurance agents to complete insurance business communication with target users, collect insurance-related audio and video data in advance, and save manpower for audio and video data collection. It should be noted that the data collection and processing before the insurance proposal is generated is mainly achieved through the front prompt generation module, specifically, the target user's basic information, insurance question and answer data, and user's current behavior data are collected in advance through the prompt generation module.

[0084] It should be noted that after the user's basic information, insurance Q&A data and user's current behavior data are acquired, the user's basic information, insurance Q&A data and user's current behavior data are preprocessed. It should be noted that the preprocessing includes: removing duplicate content, filling missing values, formatting dates and formatting data categories. The preprocessing can ensure the quality and consistency of the collected user's basic information, insurance Q&A data and user's current behavior data, and facilitate the subsequent preference prediction and insurance proposal generation operations.

[0085] See also Figure 2 In some embodiments, step S102 may include but is not limited to steps S201 to S203:

[0086] Step S201, extracting features from insurance question and answer data to obtain insurance question and answer features;

[0087] Step S202, extracting features from the user's current behavior data to obtain the user's current behavior features;

[0088] Step S203, performing preference prediction through a preset preference prediction model, insurance question-and-answer features, and user current behavior features to obtain insurance preference prediction data.

[0089] In some embodiments, in steps S201 to S202, the insurance question and answer features in the insurance question and answer data are extracted, and the user current behavior features of the user current behavior data are extracted, and the behavior and question and answer situation of the target user are represented by the insurance question and answer features and the user current behavior features. Specifically, the insurance question and answer features may include the number of questions and answers, question categories, and question and answer emotions, etc., and the user current behavior features may include the number of user browsing times, user browsing insurance information, user online time, etc. This embodiment does not specifically limit the content contained in the insurance question and answer features and the user current behavior features.

[0090] In step S203 of some embodiments, the preference prediction model is trained in advance and can predict the insurance preference data, specifically predicting the preferred insurance types and the degree of preference for each insurance type. Therefore, the preference prediction is performed by inputting the insurance question and answer features and the user's current behavior features into the preference prediction model to output the insurance preference data, and the insurance preference data is the target user's preferred insurance types and the preference probability of each preferred insurance type. The target user's preferred insurance types and preference probabilities can be used to determine the insurance types that the target user is interested in, so that the insurance proposal prepared based on the insurance types that the target user is interested in can improve the target user's experience.

[0091] In steps S201 to S203 shown in this embodiment, by extracting insurance question and answer features and user current behavior features, and inputting the insurance question and answer features and user current behavior features into the preference prediction model for preference prediction, the prediction operation of the type of insurance that the target user is interested in is made simple and accurate, so as to facilitate the preparation of insurance recommendations that meet the preferences of the target user based on the predicted insurance types.

[0092] In step S103 of some embodiments, in order to generate a more professional insurance proposal that meets the requirements of the target user, historical insurance proposals, current insurance product information and insurance claims data are obtained in advance on the insurance business platform, and the historical insurance proposals mainly include insurance proposals manually uploaded by insurance agents and insurance proposals historically generated by the insurance business platform, so the historical insurance proposals have a certain reference value when automatically generating insurance proposals. The current insurance product information is the insurance product information that can be purchased on the insurance business platform, and the current insurance product information is stored in the insurance product library. It should be noted that the insurance product library will be updated regularly, and the insurance product information that has been removed from the shelves will be removed from the insurance product library regularly, and the newly-listed insurance product information will be stored in the insurance product library to reduce the situation where the insurance product information that has been removed from the shelves is used as the current insurance product information. The insurance claims data is mainly the insurance claims plan, which consists of the insurance type category, risk category, compensation amount, case introduction and comment information. The insurance claims plan can be used to understand the claims of different insurance products under different circumstances, so that the target user can have a better understanding of the claims of insurance products.

[0093] As disclosed above, after obtaining historical insurance proposals, current insurance product information, and insurance claims data, it is necessary to extract content that can construct a professional insurance proposal from the historical insurance proposals, current insurance product information, and insurance claims data as the material for constructing the insurance proposal. Therefore, keyword extraction rules are set in advance, and the keyword extraction rules represent the key fields that need to be extracted as operation instructions for key field extraction. Among them, the keyword extraction rules include: field splitting rules and keyword classification rules, the field splitting rules serve as operation instructions for splitting historical insurance proposals, current insurance product information, and insurance claims data, and the keyword classification rules serve as operation instructions for classifying the fields after splitting. Therefore, by setting the field splitting rules and keyword classification rules, it can be determined how to complete the key field extraction of historical insurance proposals, current insurance product information, and insurance claims data.

[0094] See also Figure 3 In some embodiments, step S104 may include but is not limited to steps S301 to S303:

[0095] Step S301, performing field splitting on historical insurance proposals, current insurance product information and insurance claim data according to field splitting rules to obtain candidate fields;

[0096] Step S302, classifying each candidate field to obtain a field class;

[0097] Step S303, constructing the candidate fields into key fields according to the key field classification rules and field categories to obtain candidate insurance key fields.

[0098] In step S301 of some embodiments, historical insurance proposals, current insurance product information, and insurance claims data are split separately according to field splitting rules. Specifically, historical insurance proposals, current insurance product information, and insurance claims data are split separately according to field splitting rules such as entities, attributes, and relationships to obtain entity fields, attribute fields, and relationship fields, and the entity fields, attribute fields, and relationship fields are combined into candidate fields.

[0099] Specifically, entity fields include: user fields, insurance product fields, disease fields, medical facility fields, and insurance clause fields, etc., attribute fields are the attributes of entities in historical insurance proposals, current insurance product information, and insurance claims data, and attribute fields include: user age field, gender field, occupation field, insurance product name field, category field, premium field, and insured amount field, etc.; relationship fields are the relationships between entity fields, such as the purchase relationship between users and insurance products, the coverage relationship between diseases and insurance products, etc. Therefore, by extracting entity fields, attribute fields, and relationship fields from historical insurance proposals, current insurance product information, and insurance claims data, various insurance knowledge in the insurance field can be clearly understood.

[0100] In steps S302 to S303 of some embodiments, in order to facilitate the subsequent screening of key fields, it is necessary to classify the candidate fields into key content categories, also known as knowledge point division. Specifically, the field category of the candidate field is first identified, and the field category mainly characterizes which knowledge category in the insurance field the candidate field involves. Among them, the knowledge categories divided in the insurance field mainly include: product knowledge category, user knowledge category, disease knowledge category, claims knowledge category and question and answer knowledge category. Therefore, the candidate fields are divided according to the field category and keyword classification rules to obtain candidate insurance key fields, so the candidate insurance key fields are key fields divided according to the field category. Specifically, the candidate fields are divided into different preset knowledge bases according to the insurance category to obtain the target knowledge base, and then the fields stored in each target knowledge base are used as candidate insurance key fields.

[0101] It should be noted that the target knowledge base includes: product knowledge base, user knowledge base, disease knowledge base, claims knowledge base and question-and-answer knowledge base, and the product knowledge base covers candidate fields for all insurance products, and the corresponding content of the candidate fields includes: product name, product category, premium, insured amount, coverage and terms, etc. The user knowledge base contains candidate fields corresponding to user information, and the corresponding content of the candidate fields includes: user's age, gender, occupation and health history, etc. The disease knowledge base includes candidate fields related to diseases, and the corresponding content of the candidate fields includes: disease name, definition, common treatment methods, related insurance products, etc. The claims knowledge base includes candidate fields related to claims, and the corresponding content of the candidate fields includes: claims process, claims materials and claims introduction, etc. The question-and-answer knowledge base includes candidate fields related to questions and answers, and the corresponding content of the candidate fields includes: question-and-answer questions and answers, such as insurance, claims and product comparison, etc.

[0102] In step S301 to step S303 shown in this embodiment, when constructing the materials required for generating the insurance proposal, the existing historical insurance proposal, current insurance product information and insurance claim data are first split into candidate fields, and then the candidate fields are classified according to the key field classification rules and the field categories of the candidate fields to obtain the candidate insurance key fields. Therefore, the fields are first split and then classified, making the generation operation of the candidate insurance key fields easier and more accurate.

[0103] See also Figure 4 In some embodiments, step S105 may include but is not limited to steps S401 to S402:

[0104] Step S401, obtaining field label information of candidate insurance key fields; wherein the field label information includes: candidate user information and candidate insurance attribute information;

[0105] Step S402, screening candidate insurance key fields according to user basic information, insurance preference prediction data, candidate user information and candidate insurance attribute information to obtain target insurance key fields.

[0106] In step S401 of some embodiments, field label information is set for each candidate insurance key field, and the field label information serves as reference indicator information for screening the candidate insurance key fields. Among them, the field label information includes candidate user information and candidate insurance attribute information, and the candidate user information represents the user information associated with the candidate insurance key field, and the candidate insurance attribute information represents the insurance attribute information associated with the candidate insurance key field. It should be noted that the candidate user information specifically includes: the user's age, gender, occupation, disease information, and tolerable risk information, etc., and the candidate insurance attribute information includes: the name and category of the insurance, applicable group information, applicable disease information, etc. Therefore, the user information and insurance type that match the candidate insurance key field can be determined through the candidate user information and the candidate insurance attribute information.

[0107] In step S402 of some embodiments, the target insurance key fields are filtered out from the candidate insurance key fields in combination with the user basic information, insurance preference prediction data, candidate user information and candidate insurance attribute information. Then, the filtered out target insurance key fields are consistent with the basic information and insurance preferences of the target user, so the target insurance proposal constructed through the target insurance key fields can also improve the satisfaction of the target user.

[0108] In step S401 to step S402 shown in this embodiment, the candidate user information and the candidate insurance attribute information of the candidate insurance key fields are obtained. Then, the target insurance key fields are screened out from the candidate insurance key fields in combination with the candidate user information, the candidate insurance attribute information, the user basic information and the insurance preference prediction data, so as to accurately screen out the target insurance key fields that meet the basic information and preferences of the target user, and construct a target insurance proposal that meets the preferences and basic conditions of the target user.

[0109] See also Figure 5 In some embodiments, step S402 may include but is not limited to steps S501 to S504:

[0110] Step S501, converting the user basic information and the insurance preference prediction data into vectors to obtain a query vector;

[0111] Step S502, performing vector conversion on the candidate user information and the candidate insurance attribute information to obtain a reference vector;

[0112] Step S503, calculating the similarity between the query vector and the reference vector to obtain vector similarity;

[0113] Step S504, screening the candidate insurance key fields according to the vector similarity to obtain the target insurance key fields.

[0114] In step S501 of some embodiments, the user basic features of the user basic information are extracted, and the insurance preference features of the insurance preference prediction data are extracted, and the user basic features and the insurance preference features are spliced ​​into user query features. Then, the user query features are encoded by an encoder to obtain a query vector, which represents the user basic features and the insurance preference features in the form of a vector, so as to facilitate the screening of candidate insurance key fields according to the query vector.

[0115] In step S502 of some embodiments, candidate user features of the candidate user information are extracted, and then candidate insurance attribute features of the candidate insurance attribute information are extracted, and then the candidate user features and the candidate insurance attribute features are spliced ​​into reference features. The reference features are encoded by an encoder to obtain a reference vector represented by a vector.

[0116] In step S503 of some embodiments, the query vector and the reference vector are similarly calculated, and the similarity calculation method can be implemented by using a vector similarity function. In this embodiment, the vector similarity function includes any one of the following: cosine similarity function, Euclidean distance function, Manhattan distance function and Hamming distance function, and this embodiment does not impose specific restrictions on the vector similarity function. It should be noted that the vector similarity can represent the degree of matching between the candidate insurance key fields and the user's basic information and insurance preference prediction data, and the greater the vector similarity, the higher the degree of matching between the candidate insurance key fields and the user's basic information and insurance preference prediction data, and vice versa.

[0117] In step S504 of some embodiments, the candidate insurance key fields of each field category are sorted according to vector similarity, and a preset number of candidate insurance key fields before sorting of each field category are obtained as target insurance key fields to screen out target insurance key fields that highly match the preferences and basic information of the target user.

[0118] In steps S501 to S504 shown in this embodiment, the user basic information, insurance preference prediction data, candidate user information and candidate insurance attribute information are represented in vector form, and then the screening operation in the candidate insurance key fields is completed in a vector similarity manner, so that the screening operation of the target insurance key fields is simple and in line with the basic information and preferences of the target user.

[0119] See also Figure 6 In some embodiments, after step S106, the insurance proposal generation method further includes but is not limited to steps S601 to S603:

[0120] Step S601, performing a quality assessment on the target insurance proposal according to the preset verification insurance proposal and the preset quality assessment rules, and obtaining quality assessment data; wherein the quality assessment data represents the matching degree between the target insurance proposal and the verification insurance proposal, and the matching degree between the target insurance proposal and the quality assessment rules;

[0121] Step S602, optimizing the original proposal generation model according to the quality assessment data to obtain a target proposal generation model;

[0122] Step S603: Optimize the target insurance proposal through the target proposal generation model, user basic information and target insurance key fields to obtain an updated insurance proposal.

[0123] In step S601 of some embodiments, at the initial stage of insurance proposal generation, a verification insurance proposal that can verify the quality of the target insurance proposal generation needs to be set in advance. Therefore, the target insurance proposal generated for the first time will not be directly output to the target user, but needs to be sent to the insurance agent first, and then the insurance agent will verify the insurance proposal and quality assessment rules based on the feedback of the target insurance proposal. Then, the target insurance proposal is quality assessed in combination with the verification insurance proposal and the quality assessment rules to determine the quality assessment data of the target insurance proposal. It should be noted that the quality assessment data represents the content quality of the target insurance proposal, and the content quality includes: completeness, clarity, insurance type matching degree, product rationality, readability, compliance, etc. Specifically, the higher the quality assessment data, the better the content quality of the target insurance proposal, and conversely, the lower the quality assessment data, the worse the content quality of the target insurance proposal.

[0124] It should be noted that the original proposal generation model is trained by a large language model with the ability to generate proposals, and the prompt generation module will input the collected user basic information and the screened target insurance key fields into the original proposal generation model for proposal generation.

[0125] In step S602 of some embodiments, the quality assessment data can also characterize the generation effect of the original proposal generation model, so the model parameters of the original proposal generation model are adjusted according to the quality assessment data, and the quality assessment data is continuously updated and the original proposal generation model is optimized until the updated quality assessment data converges, and the corresponding original proposal generation model is used as the target proposal generation model.

[0126] In step S603 of some embodiments, an updated insurance proposal is generated through a target proposal generation model, basic user information, and target insurance key fields, the target insurance proposal is replaced with the updated insurance proposal, and the updated insurance proposal is sent to the target user, thereby achieving high-quality and automated insurance proposal generation.

[0127] In steps S601 to S603 shown in this embodiment, the quality of the target insurance proposal is evaluated according to the insurance proposal verification and quality evaluation rules to obtain quality evaluation data, and the original proposal generation model is adjusted according to the quality evaluation data to obtain the target proposal generation model, and then the optimized updated insurance proposal is regenerated through the target proposal generation model, user basic information and target insurance key fields to obtain an insurance proposal with better quality.

[0128] See also Figure 7 In some embodiments, step S601 may include but is not limited to steps S701 to S704:

[0129] Step S701, calculating the matching degree based on the verification insurance proposal and the target insurance proposal to obtain the content matching degree;

[0130] Step S702, extracting evaluation reference indicator data from the target insurance proposal; wherein the evaluation reference indicator data includes: insurance product information, insurance service information and insured user information;

[0131] Step S703, calculating the matching degree of the insurance product information, the insurance service information and the insured user information according to the quality assessment rule to obtain the information matching degree;

[0132] Step S704: concatenate the content matching degree and the information matching degree to obtain quality evaluation data.

[0133] It should be noted that when evaluating the content quality of the target insurance proposal, the content quality assessment can be completed from aspects such as the target insurance proposal's demand matching degree, product rationality, plan completeness, information accuracy, content readability, content clarity, content compliance, content practicality and subsequent serviceability, so as to achieve a more refined quality assessment.

[0134] In step S701 of some embodiments, as disclosed above, evaluating the content quality of the target insurance proposal involves the degree of matching, so the content features of the verification insurance proposal are extracted as the verification content features, and then the content features of the target insurance proposal are extracted as the target content features, and the content matching degree is obtained by calculating the matching degree between the verification content features and the target content features. It should be noted that the content matching degree can represent the similarity between the target insurance proposal and the verification insurance proposal, and the professionalism of the target insurance proposal.

[0135] In steps S702 to S703 of some embodiments, in order to further evaluate the product adaptation degree, service adaptation degree and user information accuracy in the target insurance proposal, insurance product information, insurance service information and insured user information are extracted from the target insurance proposal in advance, and the quality assessment rules include verifying product information, verifying service information and verifying user information. The product matching degree is obtained by matching the insurance product information and the verified product information, the service matching degree is obtained by matching the insurance service information and the verified service information, and the user matching degree is obtained by matching the insured user information and the verified user information. Then, the product matching degree, service matching degree and user matching degree are weightedly summed to obtain the information matching degree.

[0136] It should be noted that the product matching degree indicates whether the product of the target insurance proposal meets the product needs, risk tolerance, cost tolerance, etc. of the target users; the service matching degree indicates whether the insurance service of the target insurance proposal meets the insurance service needs of the target users, such as insurance terms, compensation ratio, compensation method, protection period, claim settlement process, etc.; the user matching degree indicates the accuracy of the user information in the target insurance proposal. In addition, whether the target insurance proposal has practical content, subsequent service support instructions, etc. can also be used as additional evaluation data, and the additional evaluation data can be superimposed on the quality evaluation data.

[0137] In step S704 of some embodiments, a first weight is obtained by evaluating the importance of the content match, and a second weight is obtained by evaluating the importance of the information match. The content match and the information match are spliced ​​into quality assessment data based on the first weight and the second weight.

[0138] In steps S701 to S704 illustrated in this embodiment, the quality assessment of the target insurance proposal is completed by determining the degree of match between the target insurance proposal and the verification insurance proposal, as well as the degree of match between the insurance products, insurance services and insured users in the target insurance proposal and the quality assessment rules, so that the quality assessment of the target insurance proposal is more refined and the representation of content quality is more accurate.

[0139] Please refer to Figure 8In the process of generating insurance proposals, the original proposal generation model generates multiple target insurance proposals, and then the content quality of each target insurance proposal is evaluated to obtain quality evaluation data. Then, the multiple target insurance proposals and the quality evaluation data of the target insurance proposals are displayed together on the insurance business platform, which can be provided to the target user for selection. The quality evaluation data is represented by a score. The higher the score, the better the content quality of the target insurance proposal. For example, target insurance proposal A, target insurance proposal B, and target insurance proposal C are generated respectively. The quality evaluation data of target insurance proposal A is 85, the quality evaluation data of target insurance proposal B is 79, and the quality evaluation data of target insurance proposal C is 92.

[0140] Please refer to Fig. 9 , Fig. 9 The model structure used in the insurance proposal generation method provided by the embodiment of the present application is shown, including a prompt generation module and a large language model. The prompt generation module collects the data required for the insurance proposal in advance, and then the prompt generation module pre-processes the collected data, and collects the historical proposals, insurance product library, claim case library, etc. uploaded in advance by the insurance agent through the prompt generation module. Then the historical proposals, insurance product library and claim case library are divided into product knowledge base, user knowledge base, disease knowledge base, claim knowledge base and question-and-answer knowledge base. When generating an insurance proposal, the prompt generation module extracts knowledge base fragments that match the user portrait of the target user from multiple knowledge bases, and the user portrait records the basic information, insurance preferences and disease conditions of the target user. Then, the prompt generation module inputs the knowledge base fragments and the user portrait into the large language model together, and the large language model automatically outputs the target insurance proposal. When the target insurance proposal is generated, it will not be sent directly to the target user, but will be sent to the terminal where the insurance agent is located first, and the insurance agent will review the target insurance proposal, and send the approved target insurance proposal to the target user. Therefore, not only can the automatic generation of insurance proposals be achieved, but also insurance proposals that meet the basic information and preferences of target users can be generated, thereby improving the target users' satisfaction with the insurance proposals and thus improving the professionalism of insurance services.

[0141] See also Fig.10 The embodiment of the present application further provides an insurance proposal generation device, which can implement the above-mentioned insurance proposal generation method, and the device includes:

[0142] The user data acquisition module 1001 is used to acquire the basic information of the target user, the insurance question and answer data and the current behavior data of the user; wherein the current behavior data of the user represents the operation behavior of the target user on the insurance business platform;

[0143] The preference prediction module 1002 is used to predict the preferences of target users based on the insurance question-and-answer data and the user's current behavior data to obtain insurance preference prediction data;

[0144] The historical data acquisition module 1003 is used to acquire historical insurance proposals, current insurance product information and insurance claims data;

[0145] The key field splitting module 1004 is used to split the key fields of historical insurance proposals, current insurance product information and insurance claim data according to preset key field extraction rules to obtain candidate insurance key fields;

[0146] The key field screening module 1005 is used to screen target insurance key fields from candidate insurance key fields according to user basic information and insurance preference prediction data;

[0147] The proposal screening module 1006 is used to generate a proposal based on a preset original proposal generation model, basic user information and target insurance key fields to obtain a target insurance proposal.

[0148] The specific implementation of the insurance proposal generating device is substantially the same as the specific implementation of the above-mentioned insurance proposal generating method, and will not be described in detail herein.

[0149] The embodiment of the present application also provides a computer device, the computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned insurance proposal generation method when executing the computer program. The computer device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0150] See also Fig.11 , Fig.11 The hardware structure of a computer device according to another embodiment is shown, and the computer device includes:

[0151] The processor 1101 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0152] The memory 1102 may be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1102 may store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1102, and the processor 1101 calls and executes the insurance proposal generation method of the embodiment of this application;

[0153] Input / output interface 1103, used to implement information input and output;

[0154] The communication interface 1104 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0155] A bus 1105 that transmits information between various components of the device (e.g., the processor 1101, the memory 1102, the input / output interface 1103, and the communication interface 1104);

[0156] The processor 1101 , the memory 1102 , the input / output interface 1103 and the communication interface 1104 are connected to each other in communication within the device via the bus 1105 .

[0157] The embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned insurance proposal generation method is implemented.

[0158] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0159] The insurance proposal generation method and device, computer equipment and storage medium provided in the embodiments of the present application predict the insurance preference prediction data of the target object based on the current user behavior data and insurance question and answer data of the target user on the insurance business platform, and then split the historical insurance proposals, current insurance product information and insurance claims data on the insurance business platform into candidate insurance key fields for constructing insurance proposals in advance. Therefore, when constructing an insurance proposal, the target insurance key fields that can be used to construct the insurance proposal are screened from the candidate insurance key fields based on the user's basic information and insurance preference prediction data, and the target insurance key fields are combined into the target insurance proposal using the original proposal generation model and the user's basic information. Therefore, by automatically generating insurance proposals, and the generation process will be constructed based on the basic information and insurance preferences of the target user, not only the professional requirements and manpower for the generation of insurance proposals are reduced, but also insurance proposals that meet the target user can be generated, thereby improving the quality of insurance services.

[0160] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0161] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0162] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0163] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0164] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar users, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0165] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated users, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previously associated users are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0166] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0167] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0168] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0169] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes multiple instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store programs.

[0170] The preferred embodiments of the present application are described above with reference to the accompanying drawings, but the scope of the rights of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present application should be within the scope of the rights of the present application.

Claims

1. A method for generating an insurance proposal, characterized in that: The method comprises: Obtaining basic user information, insurance question and answer data, and current user behavior data of the target user; wherein the current user behavior data represents the operation behavior of the target user on the insurance business platform; Predicting the preferences of the target user based on the insurance question-and-answer data and the user's current behavior data to obtain insurance preference prediction data; Access historical insurance proposals, current insurance product information, and insurance claims data; Extract key fields from the historical insurance proposal, the current insurance product information and the insurance claim data according to preset key field extraction rules to obtain candidate insurance key fields; wherein the candidate insurance key fields are key contents for constructing the insurance proposal; Filtering target insurance key fields from the candidate insurance key fields according to the user basic information and the insurance preference prediction data; The target insurance proposal is obtained by generating a proposal based on a preset original proposal generation model, the user basic information and the target insurance key fields.

2. The method according to claim 1, characterized in that The key field extraction rules include: field splitting rules and key field classification rules; The key field extraction rules are used to extract the key fields of the historical insurance proposal, the current insurance product information and the insurance claim data to obtain candidate insurance key fields, including: Perform field splitting on the historical insurance proposal, the current insurance product information and the insurance claim data according to the field splitting rule to obtain candidate fields; Performing category identification on each of the candidate fields to obtain a field category; The candidate fields are constructed into key fields according to the key field classification rules and the field categories to obtain the candidate insurance key fields.

3. The method according to claim 1, characterized in that The step of selecting a target insurance key field from the candidate insurance key fields according to the user basic information and the insurance preference prediction data includes: Acquire field tag information of the candidate insurance key field; wherein the field tag information includes: candidate user information and candidate insurance attribute information; The candidate insurance key fields are screened according to the user basic information, the insurance preference prediction data, the candidate user information and the candidate insurance attribute information to obtain the target insurance key fields.

4. The method according to claim 3, characterized in that The screening process of the candidate insurance key fields according to the user basic information, the insurance preference prediction data, the candidate user information and the candidate insurance attribute information to obtain the target insurance key fields includes: Performing vector conversion on the user basic information and the insurance preference prediction data to obtain a query vector; Performing vector conversion on the candidate user information and the candidate insurance attribute information to obtain a reference vector; Calculating similarity between the query vector and the reference vector to obtain vector similarity; The candidate insurance key fields are screened according to the vector similarity to obtain the target insurance key fields.

5. The method according to any one of claims 1 to 4, characterized in that: The step of predicting the preference of the target user based on the insurance question-and-answer data and the user's current behavior data to obtain insurance preference prediction data includes: Extracting features from the insurance question and answer data to obtain insurance question and answer features; Extracting features from the user's current behavior data to obtain features of the user's current behavior; Preference prediction is performed through a preset preference prediction model, the insurance question and answer characteristics, and the user's current behavior characteristics to obtain the insurance preference prediction data.

6. The method according to any one of claims 1 to 4, characterized in that: After generating a proposal through the preset original proposal generation model, the user basic information and the target insurance key fields to obtain the target insurance proposal, the method further includes: Performing a quality assessment on the target insurance proposal according to a preset verification insurance proposal and a preset quality assessment rule to obtain quality assessment data; wherein the quality assessment data represents the degree of matching between the target insurance proposal and the verification insurance proposal, and the degree of matching between the target insurance proposal and the quality assessment rule; Optimizing the original proposal generation model according to the quality assessment data to obtain a target proposal generation model; The target insurance proposal is optimized through the target proposal generation model, the user basic information and the target insurance key fields to obtain an updated insurance proposal.

7. The method according to claim 6, characterized in that The step of performing a quality assessment on the target insurance proposal according to the preset verification insurance proposal and the preset quality assessment rules to obtain quality assessment data includes: Calculate the matching degree based on the verification insurance proposal and the target insurance proposal to obtain the content matching degree; Extracting evaluation reference indicator data from the target insurance proposal; wherein the evaluation reference indicator data includes: insurance product information, insurance service information and insured user information; Calculating the matching degree of the insurance product information, the insurance service information and the insured user information according to the quality assessment rule to obtain the information matching degree; The content matching degree and the information matching degree are concatenated to obtain the quality assessment data.

8. An insurance proposal generating device, characterized in that: The device comprises: A user data acquisition module is used to acquire basic user information, insurance question and answer data, and user current behavior data of a target user; wherein the user current behavior data represents the operation behavior of the target user on the insurance business platform; A preference prediction module, used to predict the preference of the target user based on the insurance question-and-answer data and the user's current behavior data, to obtain insurance preference prediction data; The historical data acquisition module is used to obtain historical insurance proposals, current insurance product information and insurance claims data; A key field splitting module is used to split the historical insurance proposal, the current insurance product information and the insurance claim data into key fields according to a preset key field extraction rule to obtain candidate insurance key fields; A key field screening module, used for screening target insurance key fields from the candidate insurance key fields according to the user basic information and the insurance preference prediction data; The proposal screening module is used to generate a proposal through a preset original proposal generation model, the user basic information and the target insurance key fields to obtain a target insurance proposal.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the insurance proposal generation method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the insurance proposal generating method according to any one of claims 1 to 7 is implemented.

Images