Insurance product generation method and device, equipment and medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an insurance product generation method and device, equipment and a medium, and the method comprises the steps: collecting health data and behavior mode data of a target user, carrying out the feature extraction of the health data and behavior mode data, and obtaining a structured feature vector; analyzing the structured feature vector through an artificial intelligence model to obtain a plurality of candidate insurance products; inputting the plurality of candidate insurance products into an insurance product evaluation model, so that the insurance product evaluation model scores the plurality of candidate insurance products based on the historical claim settlement data and the user feature matching degree, and obtaining an insurance score value corresponding to each candidate insurance product; and determining the candidate insurance product with the highest insurance score value as an optimal insurance product, and constructing an insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base. The method can be applied to a financial science and technology business program system, and the insurance product text which is complete in structure and meets user requirements can be constructed.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, equipment, and medium for generating insurance products. Background Technology

[0002] In the financial insurance sector, with the development of big data and artificial intelligence technologies, users' demands for personalized and intelligent insurance products are constantly growing. Traditional insurance product design typically relies on fixed templates and manual rule configuration, making it difficult to accurately match users' health conditions, lifestyles, and risk preferences, resulting in severe product homogenization and insufficient user stickiness. Simultaneously, insurance companies lack a deep understanding of user behavior data during product development, leading to risk assessments deviating from actual underwriting needs. Furthermore, existing product evaluation systems are mostly based on static analysis, failing to fully utilize the dynamic correlation between historical claims data and individual profiles, thus affecting the scientific nature and profitability of insurance product configuration.

[0003] To address the aforementioned issues, there is an urgent need for a method to generate insurance products that can dynamically generate candidate products based on user characteristics, thereby constructing a complete insurance product text that meets user needs. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for generating insurance products, in order to solve the technical problem in related technologies that it is impossible to construct insurance products with complete structure and that meet user needs.

[0005] Firstly, a method for generating an insurance product is provided, the method comprising:

[0006] Collect health data and behavioral pattern data of target users, and perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors;

[0007] The structured feature vectors are analyzed using an artificial intelligence model to obtain multiple candidate insurance products;

[0008] The multiple candidate insurance products are input into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product;

[0009] The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and the corresponding insurance product text is constructed based on a preset insurance knowledge base.

[0010] Secondly, an apparatus for generating an insurance product is provided, comprising:

[0011] The data acquisition module is used to collect health data and behavioral pattern data of the target user, and to perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors.

[0012] The analysis module is used to analyze the structured feature vectors using an artificial intelligence model to obtain multiple candidate insurance products;

[0013] The scoring module is used to input the multiple candidate insurance products into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product;

[0014] The generation module is used to determine the candidate insurance product with the highest insurance score as the optimal insurance product, and to construct the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base.

[0015] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for generating the insurance product.

[0016] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described method for generating an insurance product.

[0017] The solution implemented by the aforementioned insurance product generation method, apparatus, computer equipment, and storage medium includes the following steps: collecting health data and behavioral pattern data of target users, and performing feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors; analyzing the structured feature vectors through an artificial intelligence model to obtain multiple candidate insurance products; inputting the multiple candidate insurance products into an insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree between historical claims data and user characteristics, obtaining an insurance score value corresponding to each candidate insurance product; determining the candidate insurance product with the highest insurance score value as the optimal insurance product, and constructing the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base. In this invention, by collecting user health data and behavioral patterns, and using an artificial intelligence model to complete structured feature extraction and analysis, it is possible to accurately characterize users' health risks and consumption preferences, providing strong data support for insurance companies to achieve personalized product design. Compared with the traditional method of relying on human experience to formulate insurance plans, this method can automatically generate multiple candidate insurance products and intelligently score them based on the matching degree between historical claims data and user characteristics, thereby selecting the optimal product and improving the fit between insurance products and user needs. This process not only improves customer satisfaction and renewal rates for financial institutions, but also enhances risk control capabilities and the scientific basis of pricing. Ultimately, standardized insurance texts are generated through a knowledge base, achieving full automation from design to delivery, effectively reducing labor costs and operational risks. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of an application environment for a method of generating insurance products according to an embodiment of the present invention;

[0020] Figure 2 This is a flowchart illustrating a method for generating an insurance product according to an embodiment of the present invention;

[0021] Figure 3 yes Figure 2 A schematic diagram of a specific implementation method for step S10;

[0022] Figure 4 This is a schematic diagram of the structure of an insurance product generation device according to an embodiment of the present invention;

[0023] Figure 5This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention;

[0024] Figure 6 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] The method for generating insurance products provided in this embodiment of the invention can be applied to, for example... Figure 1 In this application environment, the client communicates with the server via a network. The server can collect health data and behavioral pattern data of the target user through the client, and perform feature extraction on the health data and behavioral pattern data to obtain a structured feature vector. The structured feature vector is then analyzed using an artificial intelligence model to obtain multiple candidate insurance products. These candidate insurance products are input into an insurance product evaluation model, which scores the candidate insurance products based on historical claims data and user feature matching, obtaining an insurance score value for each candidate insurance product. The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and an insurance product text corresponding to the optimal insurance product is constructed based on a preset insurance knowledge base. The insurance product text is then fed back to the client. In this invention, by collecting user health data and behavioral patterns and using an artificial intelligence model to complete structured feature extraction and analysis, it is possible to accurately characterize users' health risks and consumption preferences, providing strong data support for insurance companies to achieve personalized product design. Compared to traditional methods that rely on human experience to develop insurance plans, this method can automatically generate multiple candidate insurance products and intelligently score them based on the matching degree between historical claims data and user characteristics, thereby selecting the optimal product and improving the fit between insurance products and user needs. This process not only improves customer satisfaction and renewal rates for financial institutions but also enhances risk control capabilities and the scientific nature of pricing. Finally, standardized insurance text is generated through a knowledge base, achieving full automation from design to delivery, effectively reducing labor costs and operational risks. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a dedicated server or a server cluster consisting of multiple servers. The invention will be described in detail below through specific embodiments.

[0027] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a method for generating an insurance product according to an embodiment of the present invention, the method comprising the following steps:

[0028] S10: Collect the target user's health data and behavior pattern data, and perform feature extraction operations on the health data and behavior pattern data to obtain a structured feature vector.

[0029] For example, in step S10, health data (such as weight, blood pressure, and medical history) and behavioral pattern data (such as exercise frequency, dietary habits, and sleep patterns) of the target user can be collected. Data processing algorithms are used to clean, standardize, and normalize the raw data. Furthermore, artificial intelligence technologies (such as feature engineering or deep learning feature extraction) are used to extract key features reflecting the user's health status and lifestyle. Finally, these features are represented in the form of structured feature vectors, providing a quantitative basis for subsequent intelligent analysis and precise matching of insurance products, which helps improve the accuracy of risk identification and the scientific nature of insurance pricing.

[0030] Among them, such as Figure 3 As shown, step S10, which involves performing feature extraction on the health data and the behavioral pattern data to obtain a structured feature vector, includes the following steps:

[0031] S11: Perform data standardization processing on the health data to obtain a health indicator dataset; and perform semantic encoding processing on the behavior pattern data to obtain a behavior pattern vector.

[0032] S12: Perform feature fusion between the health indicator dataset and the behavior pattern vector to obtain a unified feature representation.

[0033] S13: Analyze the unified feature representation through the feature representation generation model to obtain the structured feature vector.

[0034] For example, step S11 can be used to standardize and semantically encode health data and behavioral pattern data, respectively. After standardization, health data forms a health indicator dataset, which helps eliminate the influence of different data dimensions and units, improving the accuracy of subsequent analysis. Behavioral pattern data, such as users' daily routines, exercise, diet, and work habits, is transformed into quantifiable behavioral pattern vectors using semantic encoding methods (e.g., encoding mechanisms based on word vectors or graph neural networks), ensuring that unstructured information can be effectively identified and utilized. This process is particularly suitable for the deep modeling needs of individual user characteristics in the insurance and finance sector.

[0035] Furthermore, in steps S12 and S13, the health indicator dataset and behavioral pattern vectors can be fused to generate a unified feature representation. This representation is then analyzed by a feature representation generation model to output a final structured feature vector. This feature vector not only comprehensively reflects the user's overall characteristics in terms of health risk and behavioral preferences but also exhibits good model adaptability, serving as input for subsequent artificial intelligence models to generate personalized insurance products. In financial insurance, this deep fusion and modeling approach significantly enhances the scientific rigor of product recommendations and the controllability of claims risks, providing robust intelligent support for health insurance, annuity insurance, and other insurance products.

[0036] S20: Analyze the structured feature vectors using an artificial intelligence model to obtain multiple candidate insurance products.

[0037] In some embodiments, the artificial intelligence model includes a diffusion generation module and an adaptive filtering module. The step of analyzing the structured feature vector through the artificial intelligence model to obtain multiple candidate insurance products includes: inputting the structured feature vector into the diffusion generation module in the artificial intelligence model to obtain an initial insurance product; and using the adaptive generation module in the artificial intelligence model to analyze the initial insurance product to obtain candidate insurance products that conform to preset insurance logic.

[0038] For example, in step S20, the structured feature vector, as a comprehensive characterization of the user's health status and behavioral patterns, is input into an artificial intelligence model for analysis to generate multiple candidate insurance products. The artificial intelligence model consists of a diffusion generation module and an adaptive filtering module. First, the diffusion generation module uses a diffusion-based generation strategy (e.g., based on probability diffusion or multi-path generation techniques) to model the input feature vector, generating multiple insurance product prototypes. These initial insurance products may cover various insurance combinations, coverage periods, payout structures, and premium levels, demonstrating strong diversity and coverage. This generation method can effectively explore the insurance product space, providing rich alternatives for subsequent screening.

[0039] Furthermore, the adaptive filtering module can perform in-depth analysis and optimization screening of the initially generated insurance products through an embedded rule engine or discriminative model. This module matches and eliminates products item by item based on preset insurance logic, such as regulatory compliance, the user's age range, and the insurable health condition, thereby selecting candidate insurance products that are truly suitable for the target user. In financial insurance scenarios, this collaborative mechanism of "generation and filtering" not only improves product matching efficiency but also avoids invalid recommendations and potential violations, ensuring that the final output of candidate products is accurate, compliant, and high-value, helping insurance institutions achieve personalized services and intelligent operation goals.

[0040] S30: Input the multiple candidate insurance products into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product.

[0041] In some embodiments, inputting the plurality of candidate insurance products into an insurance product evaluation model, so that the insurance product evaluation model scores the plurality of candidate insurance products based on historical claims data and user feature matching degree, to obtain an insurance score value corresponding to each candidate insurance product, includes: extracting key parameter features corresponding to each candidate insurance product; wherein, the key parameter features include coverage, claim conditions, payment period, and deductible clauses; scoring each key parameter feature based on the historical claims data and user feature matching degree through the insurance product evaluation model to obtain a historical claims score and a user matching degree score; and weighting and fusing the historical claims score and the user matching degree score to obtain the insurance score value corresponding to the candidate insurance product.

[0042] For example, in step S30, multiple candidate insurance products can be input into the insurance product evaluation model for quantitative evaluation. The core function of the insurance product evaluation model is to comprehensively evaluate the risk control capabilities and suitability of each candidate insurance product by combining historical claims data with individual user characteristics. Specifically, key parameter features can first be extracted from each candidate product, including coverage (e.g., types of diseases, payout amounts), payout conditions (e.g., waiting period, diagnostic criteria), payment cycle (e.g., monthly, annual payment), and deductible clauses. These parameters are core variables that determine the actual suitability and claims stability of insurance products, and are helpful for subsequent detailed scoring and comparative analysis.

[0043] Furthermore, the insurance product evaluation model can score the aforementioned key parameters from two dimensions: first, based on historical claims data, it measures the performance of the insurance product in actual operation, including the payout ratio, payout amount, and customer satisfaction, generating a historical payout score; second, it matches and analyzes the candidate products with the target user's feature vectors (health risk, economic capacity, behavioral patterns, etc.), generating a user matching score. These two types of scores can be integrated through a pre-set weighted fusion mechanism to obtain the final insurance score for each candidate insurance product.

[0044] In the financial insurance sector, this dual assessment mechanism based on historical data and individual matching helps to achieve rational recommendations and risk-controlled insurance purchases, improving user conversion rates while also reducing the probability of losses and operating costs for insurance companies in future claims processes.

[0045] S40: The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and the insurance product text corresponding to the optimal insurance product is constructed based on the preset insurance knowledge base.

[0046] In some embodiments, constructing the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base includes: extracting key information from the optimal insurance product to obtain core insurance elements; wherein, the core insurance elements include coverage information, sum insured setting information, claims rules information, and applicable population information; constructing the initial insurance product text corresponding to the core insurance elements based on the preset insurance knowledge base using an insurance product text generation model; and sequentially performing structural review and format standardization processing on the initial insurance product text to obtain the insurance product text.

[0047] For example, in step S40, the candidate insurance product with the highest score in the previous stage can be identified as the optimal insurance product, and the insurance product text for this product can be automatically constructed based on a preset insurance knowledge base. First, key information is extracted from the optimal product to identify its core insurance elements, including coverage information (such as disease types, hospitalization allowances, etc.), sum insured information (such as maximum payout), payout rules information (such as number of payouts, waiting period), and applicable population information (such as suitable age, gender, or health status). This step ensures that the insurance product description is highly consistent with the actual content, providing a structured basis for text construction, and is particularly suitable for scenarios where financial insurance products such as health insurance and pension insurance require strictly defined terms.

[0048] Furthermore, an insurance product text generation model, combined with a pre-set insurance knowledge base (including industry standards, legal expressions, and compliance terminology templates), can automatically organize the aforementioned core insurance elements into initial text. After generation, the text can undergo structural review (ensuring logical coherence and completeness of elements) and format standardization (e.g., segmentation, numbering, and citation according to regulatory requirements), ultimately resulting in a standardized, compliant insurance product text suitable for user display or electronic contract output. This automated process not only significantly improves the efficiency and compliance of insurance product launches but also reduces manual editing costs, which is of great significance for financial institutions to quickly respond to customer needs and flexibly design differentiated products.

[0049] In some embodiments, the method further includes: acquiring a training dataset and a pre-trained model; wherein the training dataset includes several historical candidate insurance products; labeling the training dataset to obtain labeling results, wherein the labeling results include historical insurance score values ​​corresponding to the historical candidate insurance products; and training the pre-trained model using the training dataset and the labeling results to obtain the insurance product evaluation model.

[0050] Based on the above embodiments, after obtaining the insurance product evaluation model, the method further includes: iteratively training the insurance product evaluation model based on the training dataset and the annotation results to extract data features and calculate the classification loss function; iteratively training the classification loss function using a preset method with the aim of reducing the value of the classification loss function until the value of the classification loss function is less than the expected threshold; and obtaining the iteratively trained insurance product evaluation model based on the iteratively trained classification loss function.

[0051] Specifically, a training dataset containing several historical candidate insurance products can be collected for training purposes. This dataset can be obtained through methods such as manual collection, web scraping, or publicly available datasets; this application does not limit the scope of the application.

[0052] Furthermore, each set of historical candidate insurance products can be labeled to obtain labeled results corresponding to each set of historical candidate insurance products. These labeled results are then used as labels for the input data set. Each set of labeled training datasets is then input into the pre-trained model for supervised learning. Training ends when the training termination conditions are met, such as when the number of training iterations reaches a threshold or the model's output accuracy reaches a threshold, resulting in a trained insurance product evaluation model.

[0053] In this embodiment, the training dataset and annotation results can be input into a pre-trained model for supervised learning, thereby training an insurance product evaluation model. This allows for the output of labeled results based on the insurance product evaluation model.

[0054] The above embodiments enhance data quality and diversity during the training process of insurance product evaluation models, thereby improving the model's generalization ability and practical application effectiveness.

[0055] Understandably, in order to train an insurance product evaluation model with higher accuracy, the model can be iteratively trained repeatedly to continuously reduce the classification loss function until it meets the expected threshold. In this way, a more accurate labeling result can be obtained based on the iterated insurance product evaluation model.

[0056] It should be noted that this application does not limit the above-mentioned preset method and expected threshold. For example, the preset method can be gradient descent algorithm, batch gradient descent algorithm, stochastic gradient descent algorithm, etc. This application uses gradient descent algorithm as an example for explanation.

[0057] The purpose of the gradient descent algorithm is to find the minimum value of the classification loss function, or to converge to the minimum value, through iteration. Geometrically speaking, gradient descent occurs where the gradient decreases most rapidly in the opposite direction of the vector where the function's change increases, making it easier to find the function's minimum. Based on this, in this embodiment, the gradient descent algorithm can be used to iteratively train the insurance product evaluation model, continuously reducing the classification loss function and thus decreasing the error in the calculation results.

[0058] In this embodiment, the classification loss function is continuously reduced by iteratively training the insurance product evaluation model using the gradient descent algorithm, resulting in an iterative insurance product evaluation model. Consequently, a more accurate labeling result can be obtained based on the iterative insurance product evaluation model.

[0059] As can be seen, the above solution, by collecting user health data and behavioral patterns and leveraging artificial intelligence models to extract and analyze structured features, can accurately characterize users' health risks and consumption preferences, providing strong data support for insurance companies to design personalized products. Compared to the traditional method of relying on human experience to formulate insurance plans, this method can automatically generate multiple candidate insurance products and intelligently score them based on the matching degree between historical claims data and user characteristics, thereby selecting the optimal product and improving the fit between insurance products and user needs. This process not only improves the customer satisfaction and renewal rate of financial institutions but also enhances risk control capabilities and the scientific nature of pricing. Finally, standardized insurance text is generated through a knowledge base, achieving full automation from design to delivery, effectively reducing labor costs and operational risks.

[0060] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0061] In one embodiment, an apparatus for generating insurance products is provided, which corresponds one-to-one with the method for generating insurance products in the above embodiments. For example... Figure 4 As shown, the device for generating this insurance product includes a data acquisition module 101, an analysis module 102, a scoring module 103, and a generation module 104. Detailed descriptions of each functional module are as follows:

[0062] The acquisition module 101 is used to acquire the health data and behavior pattern data of the target user, and to perform feature extraction operations on the health data and behavior pattern data to obtain a structured feature vector;

[0063] Analysis module 102 is used to analyze the structured feature vector through an artificial intelligence model to obtain multiple candidate insurance products;

[0064] The scoring module 103 is used to input the multiple candidate insurance products into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product.

[0065] The generation module 104 is used to determine the candidate insurance product with the highest insurance score as the optimal insurance product, and to construct the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base.

[0066] The acquisition module 101 is used to perform data standardization processing on the health data to obtain a health indicator dataset; and to perform semantic encoding processing on the behavior pattern data to obtain a behavior pattern vector; to perform feature fusion on the health indicator dataset and the behavior pattern vector to obtain a unified feature representation; and to analyze the unified feature representation through a feature expression generation model to obtain the structured feature vector.

[0067] Analysis module 102 is used to input the structured feature vector into the diffusion generation module in the artificial intelligence model to obtain an initial insurance product; and to analyze the initial insurance product using the adaptive generation module in the artificial intelligence model to obtain candidate insurance products that conform to preset insurance logic.

[0068] The scoring module 103 is used to extract key parameter features corresponding to each of the candidate insurance products; wherein, the key parameter features include coverage, compensation conditions, payment period, and deductible clauses; the insurance product evaluation model scores each of the key parameter features based on the historical claims data and the user feature matching degree to obtain historical claims score and user matching degree score; the historical claims score and the user matching degree score are weighted and fused to obtain the insurance score value corresponding to the candidate insurance product.

[0069] The generation module 104 is used to extract key information from the optimal insurance product to obtain core insurance elements; wherein, the core insurance elements include coverage information, sum insured setting information, claims rules information, and applicable population information; the initial text of the insurance product corresponding to the core insurance elements is constructed based on the preset insurance knowledge base through the insurance product text generation model; the initial text of the insurance product is subjected to structural review and format standardization processing in sequence to obtain the insurance product text.

[0070] In one embodiment, the acquisition module 101 is further configured to: acquire a training dataset and a pre-trained model; wherein the training dataset includes several historical candidate insurance products; label the training dataset to obtain labeling results, wherein the labeling results include historical insurance score values ​​corresponding to the historical candidate insurance products; and train the pre-trained model using the training dataset and the labeling results to obtain the insurance product evaluation model.

[0071] In one embodiment, the acquisition module 101 is further configured to: perform iterative training on the insurance product evaluation model based on the training dataset and the annotation results to extract data features and calculate the classification loss function; perform iterative training on the classification loss function using a preset method with the aim of reducing the value of the classification loss function until the value of the classification loss function is less than the expected threshold; and obtain the iterative insurance product evaluation model based on the iteratively trained classification loss function.

[0072] This invention provides an insurance product generation device. By collecting user health data and behavioral patterns, and leveraging artificial intelligence models to extract and analyze structured features, it can accurately characterize users' health risks and consumption preferences, providing strong data support for insurance companies to design personalized products. Compared to traditional methods that rely on human experience to develop insurance plans, this method can automatically generate multiple candidate insurance products and intelligently score them based on the matching degree between historical claims data and user characteristics, thereby selecting the optimal product and improving the fit between insurance products and user needs. This process not only improves customer satisfaction and renewal rates for financial institutions but also enhances risk control capabilities and the scientific nature of pricing. Finally, standardized insurance text is generated through a knowledge base, achieving full automation from design to delivery, effectively reducing labor costs and operational risks.

[0073] Specific limitations regarding the device for generating insurance products can be found in the limitations on the methods for generating insurance products described above, and will not be repeated here. Each module in the aforementioned device for generating insurance products can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0074] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a method for generating an insurance product on the server side.

[0075] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the client-side functions or steps of a method for generating an insurance product.

[0076] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0077] Collect health data and behavioral pattern data of target users, and perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors;

[0078] The structured feature vectors are analyzed using an artificial intelligence model to obtain multiple candidate insurance products;

[0079] The multiple candidate insurance products are input into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product;

[0080] The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and the corresponding insurance product text is constructed based on a preset insurance knowledge base.

[0081] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0082] Collect health data and behavioral pattern data of target users, and perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors;

[0083] The structured feature vectors are analyzed using an artificial intelligence model to obtain multiple candidate insurance products;

[0084] The multiple candidate insurance products are input into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product;

[0085] The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and the corresponding insurance product text is constructed based on a preset insurance knowledge base.

[0086] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0087] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0088] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0089] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for generating an insurance product, characterized in that, The method includes: Collect health data and behavioral pattern data of target users, and perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors; The structured feature vectors are analyzed using an artificial intelligence model to obtain multiple candidate insurance products; The multiple candidate insurance products are input into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product; The candidate insurance product with the highest insurance score is determined as the optimal insurance product, and the corresponding insurance product text is constructed based on a preset insurance knowledge base.

2. The method according to claim 1, characterized in that, The step of performing feature extraction operations on the health data and the behavioral pattern data to obtain a structured feature vector includes: The health data is standardized to obtain a health indicator dataset; and, The behavioral pattern data is semantically encoded to obtain a behavioral pattern vector; The health indicator dataset and the behavior pattern vector are fused to obtain a unified feature representation; The unified feature representation is analyzed by a feature representation generation model to obtain the structured feature vector.

3. The method according to claim 1, characterized in that, The artificial intelligence model includes a diffusion generation module and an adaptive filtering module. The structured feature vectors are analyzed using the artificial intelligence model to obtain multiple candidate insurance products, including: The structured feature vector is input into the diffusion generation module in the artificial intelligence model to obtain the initial insurance product; The initial insurance product is analyzed using the adaptive generation module in the artificial intelligence model to obtain candidate insurance products that conform to the preset insurance logic.

4. The method according to claim 1, characterized in that, The step of inputting the multiple candidate insurance products into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains an insurance score value corresponding to each candidate insurance product, includes: Extract the key parameter features corresponding to each of the candidate insurance products; wherein, the key parameter features include coverage, compensation conditions, payment period and deductible clauses; The insurance product evaluation model scores each key parameter feature based on the historical claims data and the user feature matching degree, thereby obtaining the historical claims score and the user matching degree score. The historical claims score and the user matching score are weighted and fused to obtain the insurance score value corresponding to the candidate insurance product.

5. The method according to claim 1, characterized in that, The construction of the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base includes: Key information is extracted from the optimal insurance product to obtain the core insurance elements; wherein, the core insurance elements include information on coverage liability, information on the sum insured, information on claims rules, and information on the applicable population; The initial text of the insurance product corresponding to the core insurance elements is constructed based on the preset insurance knowledge base using the insurance product text generation model. The initial text of the insurance product is sequentially subjected to structural review and format standardization to obtain the insurance product text.

6. The method according to claim 1, characterized in that, The method further includes: Obtain a training dataset and a pre-trained model; wherein the training dataset includes several historical candidate insurance products; The training dataset is labeled to obtain the labeling results, wherein the labeling results include the historical insurance score values ​​corresponding to the historical candidate insurance products; The insurance product evaluation model is obtained by training the pre-trained model using the training dataset and the annotation results.

7. The method according to claim 6, characterized in that, After obtaining the insurance product evaluation model, the process also includes: Based on the training dataset and the annotation results, the insurance product evaluation model is iteratively trained to extract data features, and the classification loss function is calculated. The classification loss function is iteratively trained using a preset method with the aim of reducing its value until the value of the classification loss function is less than the expected threshold. Based on the classification loss function after iterative training, the iterative insurance product evaluation model is obtained.

8. An apparatus for generating an insurance product, characterized in that, include: The data acquisition module is used to collect health data and behavioral pattern data of the target user, and to perform feature extraction operations on the health data and behavioral pattern data to obtain structured feature vectors. The analysis module is used to analyze the structured feature vectors using an artificial intelligence model to obtain multiple candidate insurance products; The scoring module is used to input the multiple candidate insurance products into the insurance product evaluation model, so that the insurance product evaluation model scores the multiple candidate insurance products based on the matching degree of historical claims data and user characteristics, and obtains the insurance score value corresponding to each candidate insurance product; The generation module is used to determine the candidate insurance product with the highest insurance score as the optimal insurance product, and to construct the insurance product text corresponding to the optimal insurance product based on a preset insurance knowledge base.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for generating an insurance product as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method for generating an insurance product as described in any one of claims 1 to 7.

Images