Insurance clause automatic generation method and device, equipment and storage medium
Through the method of automatically generating insurance clauses, using a large language model and a logical expression library, the problems of low efficiency in manual writing and time-consuming compliance review are solved, and efficient and accurate insurance clause design and compliance review are achieved, which has improved the Market adaptability and customer satisfaction.
Patent Information
- Application Number
- CN202411210636.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-03
AI Technical Summary
The design and generation of insurance terms relies on handwriting, is inefficient and susceptible to human factors, is difficult to respond quickly to market changes and customer needs, and compliance reviews are time-consuming and error-prone, making it impossible to efficiently ensure the legality and compliance of terms.
By obtaining insurance user feedback information and market research data for analysis, structured data analysis conclusions are generated, and prompt words are input into large language models to generate target insurance clauses, covering the scope of protection, disclaimer clauses and compensation standards, and using logical expressions to construct rules The library conducts compliance reviews and amendments to ensure compliance and market adaptability of terms.
It improves the design efficiency and compliance of insurance terms, can respond more flexibly to market demand, ensures the accuracy and consistency of terms, reduces human errors and legal risks, and improves the competitiveness and customer satisfaction of insurance products.
Smart Images

Figure CN119067786A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of insurance technology, and in particular to a method, device, equipment and storage medium for automatically generating insurance clauses. Background Art
[0002] In the insurance industry, the design and generation of insurance clauses has always been a complex and highly specialized process. Traditional insurance clause generation often relies on the experience of experts and manual writing, which is not only inefficient but also easily affected by human factors, resulting in inconsistency and inaccuracy in clauses. In addition, with the rapid changes in market demand and the diversification of customer needs, insurance companies face huge challenges and need to be able to respond to these changes flexibly and efficiently. However, the existing manual generation process can hardly meet this demand, resulting in delayed updates of insurance products and difficulty in quickly adapting to the market.
[0003] In the process of formulating insurance clauses, there is also the problem of how to ensure the compliance of the clauses. The insurance industry is subject to strict legal and regulatory requirements, and the compliance of insurance clauses is crucial to preventing legal risks. However, the process of manually writing and reviewing clauses is time-consuming and error-prone, and cannot efficiently ensure the legal compliance of the clauses.
[0004] In order to solve these problems, in recent years, the field of insurance technology (InsurTech) has begun to introduce automation technology, using big data analysis, natural language processing (NLP) and large language model (LLM) and other technical means to optimize and automate the generation and compliance review process of insurance clauses. This technological innovation not only improves the efficiency of generating insurance clauses and reduces human errors, but also enables insurance products to adapt to market demand more quickly while ensuring the compliance of clauses. Summary of the invention
[0005] The main purpose of the present invention is to provide a method, device, equipment and storage medium for automatically generating insurance clauses, aiming to solve the technical problem of how to efficiently and accurately generate insurance clauses through automated means.
[0006] To achieve the above object, the present invention provides a method for automatically generating insurance clauses, the method comprising the following steps:
[0007] Obtain feedback information from insurance users and insurance project market research data and conduct data analysis to generate structured insurance data analysis conclusions;
[0008] Obtaining existing insurance terms and conditions, and forming prompt words according to the existing insurance terms and conditions and the insurance data analysis conclusion;
[0009] The prompt words are input into a large language model to generate target insurance clauses, wherein the contents of the target insurance clauses include coverage, exemption clauses and compensation standards.
[0010] Optionally, after generating the target insurance clause, it also includes:
[0011] Collect insurance laws and regulations and insurance industry standards, and convert the insurance laws and regulations and insurance industry standards into logical expressions according to preset rules;
[0012] Insurance rules are formed according to the logical expressions, an insurance rule base is constructed, and the insurance rule base is used to conduct compliance review on the target insurance clauses.
[0013] Optionally, after using the insurance rule base to conduct compliance review on the target insurance clause, the method further includes:
[0014] A compliance report is generated for the target insurance clause. When the content of the compliance report is non-compliant, the compliance report includes a suggestion for amending the target insurance clause.
[0015] Optionally, after generating a compliance report for the target insurance clause, the method further includes:
[0016] When the content of the compliance report is non-compliant, the target insurance clause is amended according to the amendment suggestion to obtain a new version of the insurance clause, and the amendment record of the new version of the insurance clause is retained.
[0017] Optionally, after obtaining the new version of the insurance terms, it also includes:
[0018] The overall risk level of the target insurance clause is evaluated. When the overall risk level is greater than a preset threshold, the target insurance clause, amendment suggestions and the new version of the insurance clause are sent to the reviewer for review.
[0019] Optionally, after generating a compliance report for the target insurance clause, the method further includes:
[0020] When the content of the compliance report is compliant, the target insurance clause is sent to the business personnel.
[0021] Optionally, after sending the target insurance clause to the business personnel, the method further includes:
[0022] At each preset time period, the target insurance clause is evaluated using the large language model, and the content of the target insurance clause is adjusted according to the evaluation result.
[0023] Furthermore, in order to achieve the above-mentioned purpose, the present invention also provides an automatic generation device for insurance clauses, the automatic generation device for insurance clauses comprising:
[0024] The data collection and analysis module obtains feedback information from insurance users and market research data of insurance projects and performs data analysis to generate structured insurance data analysis conclusions;
[0025] A prompt word generation module obtains existing insurance terms and forms prompt words according to the existing insurance terms and the insurance data analysis conclusion;
[0026] The insurance clause generation module inputs the prompt words into the large language model to generate target insurance clauses, wherein the content of the target insurance clauses includes the coverage, exemption clauses and compensation standards.
[0027] Furthermore, to achieve the above-mentioned purpose, the present invention also provides an insurance clause automatic generation device, which includes a memory, a processor, and an insurance clause automatic generation program stored in the memory and executable on the processor, and when the insurance clause automatic generation program is executed by the processor, the steps of the insurance clause automatic generation method as described above are implemented.
[0028] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer storage medium, on which a program for automatically generating insurance terms is stored, and when the program for automatically generating insurance terms is executed by a processor, the steps of the method for automatically generating insurance terms as described above are implemented.
[0029] The present invention relates to a method for automatically generating insurance clauses, which aims to solve the problem of how to efficiently and accurately generate insurance clauses through automated means. First, by collecting feedback information from insurance users and market research data on insurance projects, these data are deeply analyzed to generate structured insurance data analysis conclusions. These conclusions provide a reliable market basis for the subsequent generation of clauses. Next, the existing insurance clauses and the generated data analysis conclusions are used to form prompt words, which will be used as input, and the large language model will generate target insurance clauses based on the prompt words. The generated clauses cover the scope of protection, exemption clauses and compensation standards, ensuring the comprehensiveness of the clauses and the market adaptability. The present invention improves the design efficiency and compliance of insurance clauses through intelligent means, and provides technical support for the automatic customization of insurance products. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 A schematic diagram of the structure of the equipment hardware operating environment involved in the embodiment of the equipment for automatically generating insurance clauses of the present invention;
[0031] Figure 2This is a flow chart of a first embodiment of the method for automatically generating insurance clauses of the present invention;
[0032] Figure 3 This is a schematic diagram of the functional modules of a preferred embodiment of the device for automatically generating insurance clauses of the present invention.
[0033] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0034] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0035] It should be noted that in the insurance industry, the design and generation of clauses has always been a complex and highly specialized process. Traditional insurance clause generation often relies on the experience and manual writing of experts, which is not only inefficient but also easily affected by human factors, resulting in inconsistency and inaccuracy of clauses. In addition, with the rapid changes in market demand and the diversification of customer needs, insurance companies face huge challenges and need to be able to respond to these changes flexibly and efficiently. However, the existing manual generation process can hardly meet this demand, resulting in lagging insurance product updates and difficulty in quickly adapting to the market. These data are the basis for enterprises to make important business decisions. The key function of the DW layer is to ensure the consistency and accuracy of data and provide reliable basic data for the data application layer.
[0036] The data application layer (DA) is the output end of the data warehouse, which extracts, filters and applies data for specific business needs. The output of the DA layer is usually used to generate reports, data analysis or other business applications. The design goal of this layer is to refine the data according to different business needs to support a variety of application scenarios.
[0037] In offline data warehouse systems, although immediate response is not required, data processing tasks must be completed within a specified time to support subsequent task chains. For example, in the regulatory reporting scenario in an investment management system, the system must complete the processing of upstream data before a specific time point and generate corresponding reports for the regulatory system to obtain and submit.
[0038] When upstream data changes, especially when processing data for a quarter or longer, the system needs to backtrack and reprocess the data for the entire time period. In this case, the performance of the data warehouse system is often severely challenged, especially when processing large amounts of historical data. Traditional data warehouse systems usually use a batch processing model to extract data from the ODS, store it in the database after processing at the DW layer, and finally extract and process the data according to demand at the DA layer. When encountering data adjustments, the system needs to backtrack and process data day by day, repeating the same data processing process. This approach often has the problem of inefficiency when dealing with large-scale data processing.
[0039] Based on the above-mentioned defects, the present invention provides an automatic generation device for insurance clauses, referring to Figure 1 , Figure 1 This is a schematic diagram of the structure of the equipment hardware operating environment involved in the embodiment of the equipment for automatically generating insurance terms of the present invention.
[0040] like Figure 1 As shown, the insurance clause automatic generation device may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.
[0041] Those skilled in the art will understand that Figure 1 The hardware structure of the device for automatically generating insurance terms shown in the figure does not constitute a limitation on the device for automatically generating insurance terms, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.
[0042] like Figure 1As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and an insurance clause automatic generation program. The operating system is a program for managing and controlling the insurance clause automatic generation device and software resources, and supports the operation of the network communication module, the user interface module, the insurance clause automatic generation program, and other programs or software; the network communication module is used to manage and control the network interface 1004; and the user interface module is used to manage and control the user interface 1003.
[0043] exist Figure 1 In the hardware structure of the device for automatically generating insurance clauses, the network interface 1004 is mainly used to connect to the backend server and perform data communication with the backend server; the user interface 1003 is mainly used to connect to the client and perform data communication with the client; the processor 1001 can call the automatic generation program of insurance clauses stored in the memory 1005 and perform the following operations:
[0044] Obtain feedback information from insurance users and insurance project market research data and conduct data analysis to generate structured insurance data analysis conclusions;
[0045] Obtaining existing insurance terms and conditions, and forming prompt words according to the existing insurance terms and conditions and the insurance data analysis conclusion;
[0046] The prompt words are input into a large language model to generate target insurance clauses, wherein the contents of the target insurance clauses include coverage, exemption clauses and compensation standards.
[0047] Furthermore, after the target insurance clause is generated, it also includes:
[0048] Collect insurance laws and regulations and insurance industry standards, and convert the insurance laws and regulations and insurance industry standards into logical expressions according to preset rules;
[0049] Insurance rules are formed according to the logical expressions, an insurance rule base is constructed, and the insurance rule base is used to conduct compliance review on the target insurance clauses.
[0050] Furthermore, after the target insurance clause is subjected to compliance review using the insurance rule base, the method further includes:
[0051] A compliance report is generated for the target insurance clause. When the content of the compliance report is non-compliant, the compliance report includes a suggestion for amending the target insurance clause.
[0052] Furthermore, after generating a compliance report for the target insurance clause, the method further includes:
[0053] When the content of the compliance report is non-compliant, the target insurance clause is amended according to the amendment suggestion to obtain a new version of the insurance clause, and the amendment record of the new version of the insurance clause is retained.
[0054] Furthermore, after obtaining the new version of the insurance terms, it also includes:
[0055] The overall risk level of the target insurance clause is evaluated. When the overall risk level is greater than a preset threshold, the target insurance clause, amendment suggestions and the new version of the insurance clause are sent to the reviewer for review.
[0056] Furthermore, after generating a compliance report for the target insurance clause, the method further includes:
[0057] When the content of the compliance report is compliant, the target insurance clause is sent to the business personnel.
[0058] Furthermore, after sending the target insurance clause to the business personnel, the method further includes:
[0059] At each preset time period, the target insurance clause is evaluated using the large language model, and the content of the target insurance clause is adjusted according to the evaluation result.
[0060] The specific implementation of the device for automatically generating insurance clauses of the present invention is basically the same as the embodiments of the method for automatically generating insurance clauses described below, and will not be described in detail here.
[0061] The present invention also provides a method for automatically generating insurance clauses based on the above-mentioned device for automatically generating insurance clauses.
[0062] Reference Figure 2 , Figure 2 This is a flow chart of the first embodiment of the method for automatically generating insurance clauses of the present invention.
[0063] The embodiment of the present invention provides an embodiment of a method for automatically generating insurance clauses. It should be noted that although a logical order is shown in the flow chart, in some cases, the steps shown or described may be executed in an order different from that shown here.
[0064] In various embodiments of the method for automatically generating insurance clauses, the execution entity is an automatic insurance clause generation device.
[0065] The method for automatically generating insurance clauses is applied to an automatic generation device for insurance clauses, and the method for automatically generating insurance clauses includes:
[0066] Step S10, obtaining feedback information from insurance users and insurance project market research data and performing data analysis to generate structured insurance data analysis conclusions;
[0067] In the insurance industry, the design and generation of clauses has always been a complex and highly specialized process. Traditional insurance clause generation often relies on the experience and manual writing of experts, which is not only inefficient but also easily affected by human factors, resulting in inconsistency and inaccuracy of clauses. In addition, with the rapid changes in market demand and the diversification of customer needs, insurance companies face huge challenges and need to be able to respond to these changes flexibly and efficiently. However, the existing manual generation process can hardly meet this demand, resulting in lagging insurance product updates and difficulty in quickly adapting to the market. These data are the basis for enterprises to make important business decisions. The key function of the DW layer is to ensure the consistency and accuracy of data and provide reliable basic data for the data application layer.
[0068] The data application layer (DA) is the output end of the data warehouse, which extracts, filters and applies data for specific business needs. The output of the DA layer is usually used to generate reports, data analysis or other business applications. The design goal of this layer is to refine the data according to different business needs to support a variety of application scenarios.
[0069] In offline data warehouse systems, although immediate response is not required, data processing tasks must be completed within a specified time to support subsequent task chains. For example, in the regulatory reporting scenario in an investment management system, the system must complete the processing of upstream data before a specific time point and generate corresponding reports for the regulatory system to obtain and submit.
[0070] When upstream data changes, especially when processing data for a quarter or longer, the system needs to backtrack and reprocess the data for the entire time period. In this case, the performance of the data warehouse system is often severely challenged, especially when processing large amounts of historical data. Traditional data warehouse systems usually use a batch processing model to extract data from the ODS, store it in the database after processing at the DW layer, and finally extract and process the data according to demand at the DA layer. When encountering data adjustments, the system needs to backtrack and process data day by day, repeating the same data processing process. This approach often has the problem of inefficiency when dealing with large-scale data processing.
[0071] In this embodiment, data analysis plays a vital role in the automatic generation of insurance clauses. Obtaining feedback information from insurance users and insurance project market research data and analyzing these data are the basic steps for generating structured insurance data analysis conclusions.
[0072] Collect feedback from insurance users from multiple channels. This feedback may include user satisfaction with existing insurance products, suggestions on coverage, expectations for compensation standards, etc. In addition, it is also necessary to collect market research data, including market demand, competitor product analysis, market trends, etc. These data provide basic information for subsequent clause generation.
[0073] The collected data is processed using big data analysis tools, including data cleaning, data organization, data mining and other steps, to extract key information related to insurance terms. The content of the analysis includes the commonalities and differences in user needs, the evolution of market trends, the advantages and disadvantages of competing product terms, etc.
[0074] The analysis results are structured to form the basis for the subsequent generation of prompts. These structured conclusions can include the core needs of users, hot product features in the market, existing terms that need to be improved, etc.
[0075] A specific implementation method may be: first, automatically obtain data from multiple data sources (such as user feedback systems, market research platforms, etc.). These data may exist in text, tables, or other formats. Then, the system processes these data through a preset analysis algorithm. For example, the system can use natural language processing (NLP) technology to perform sentiment analysis on the text content of user feedback to identify the user's satisfaction with certain terms.
[0076] Next, the system will conduct trend analysis on the market research data to identify the popular insurance products and their clause characteristics in the current market. The results of the analysis will be presented in a structured form, such as data tables, charts or analysis reports. These structured data will become the basis for generating prompt words to ensure that the generated insurance clauses can accurately reflect market demand and user preferences.
[0077] Example description:
[0078] Example 1: The system collects a large number of user reviews about "health insurance" products from the user feedback platform. After these reviews are processed by NLP, the system identifies several issues that users are generally concerned about: the limitations of coverage, the complexity of the compensation process, and the ambiguity of certain exemption clauses. The system then extracts these issues and combines them with market research data (such as competitive product analysis and market trends) to generate a structured analysis conclusion.
[0079] Example 2: In another embodiment, the system extracts several popular insurance clause features from multiple market research reports, such as "high protection, low threshold compensation". Through analysis, the system determines the popularity of these features in the target market and structures these results into an analysis report as a reference for subsequent insurance clause generation.
[0080] Through the above steps, insurance clauses that meet market and user needs can be automatically and accurately generated, thereby improving the competitiveness of insurance products and customer satisfaction.
[0081] Step S20, obtaining existing insurance terms, and forming prompt words according to the existing insurance terms and the insurance data analysis conclusion;
[0082] In this embodiment, the system extracts relevant clauses from the existing insurance clause database. These clauses include but are not limited to coverage, exemption clauses, compensation standards, etc. The extracted clause data will serve as important reference information for generating prompt words.
[0083] The system analyzes the extracted existing insurance clauses to identify the core elements, common clause structures, and possible deficiencies in the current clauses. The system compares and combines the analysis results of the existing clauses with the previously generated structured insurance data analysis conclusions. The data analysis conclusions may point out the need for improvement of existing clauses in user feedback, new trends discovered in market research, or advantageous clauses of competing products.
[0084] Based on the above analysis, the system generates prompt words for input into the large language model. These prompt words will clearly indicate the key areas that the large language model should focus on when generating new terms, such as strengthening certain guarantees, adding clear disclaimer clauses, or adjusting compensation standards to improve market competitiveness.
[0085] The generated prompt words can be in a structured format to ensure that they can be effectively processed by the large language model. For example, the prompt words can be divided into several parts, each corresponding to a specific clause type (such as coverage or compensation standards). The prompt words may also contain specific keywords or sentence patterns to ensure that the generated clauses comply with established legal and market standards.
[0086] Example description:
[0087] Example 1: The system extracts the content about "hospitalization expense coverage" from the existing health insurance clauses. Combined with the conclusion of data analysis, it is found that users generally want to increase the coverage of outpatient expenses. Therefore, the generated prompt words may be: "Generate a clause, which should increase the coverage of outpatient expenses in addition to hospitalization expense coverage, and clarify the limit and applicable conditions of the coverage."
[0088] Example 2: In another embodiment, the “disclaimer” content in the existing terms is extracted and combined with the vague description problem reported by the user, and the generated prompt is: “Ensure that the generated disclaimer contains specific disclaimer conditions and avoid using vague language to improve the clarity of the terms and user understanding.”
[0089] Through the above steps, the generation process of prompt words is based on the actual content of the existing clauses and fully considers user needs and market changes, thereby ensuring that the generated insurance clauses can better meet the needs of the market and customers.
[0090] Step S30, inputting the prompt words into a large language model to generate target insurance clauses, wherein the contents of the target insurance clauses include coverage, exemption clauses and compensation standards.
[0091] In this embodiment, the prompt words generated in the previous step are input into the large language model. The prompt words provide clear guidance on how the large language model should generate the specific content of the insurance clause. For example, the prompt words may require the model to highlight certain coverage when generating clauses, or to add detailed explanations to the exemption clauses.
[0092] The large language model parses the input prompt words to understand the intentions and requirements. By parsing the prompt words, the model determines the key points and clause structure that need to be paid attention to when generating clauses. Based on the parsing results, the model automatically generates the target insurance clauses. The model generates a complete clause text including the coverage, exemption clauses and compensation standards by integrating the requirements in the prompt words and the internal knowledge base. Each clause is refined to ensure the accuracy, compliance and readability of the language.
[0093] In order to ensure the quality of the generated clauses, the system may generate multiple rounds of clauses, each time generating a set of candidate clauses. The system will compare these candidate clauses and select the version that best meets the prompt word requirements and legal regulations. For the clauses generated for the first time, the system can also adjust them according to the internal optimization algorithm. For example, the model can fine-tune the language of the clauses according to the latest changes in laws and regulations to ensure their legality and market competitiveness.
[0094] Finally, the system outputs one or more versions of the target insurance clauses, which cover the coverage, exemption clauses and compensation standards. The output clauses can be directly applied to the insurance contract or further submitted for manual review to ensure that they fully comply with legal requirements and market needs.
[0095] Example description:
[0096] Example 1: When generating health insurance clauses, the prompt words may include: "Generate a health insurance clause for young families, focusing on covering hospitalization expenses and routine physical examination expenses, and clearly stating that the exemption clauses should include behaviors such as suicide." After entering the prompt words, the large language model automatically generates a health insurance clause, which clearly stipulates the compensation standards for hospitalization and physical examinations, and lists the exemption clauses in detail.
[0097] Example 2: In another example, the prompt word requires the generation of a travel insurance clause, which should include emergency medical care, luggage loss, flight delays, and the exemption clause should describe force majeure in detail. The large language model generates a complete travel insurance clause based on the prompt word, covering all requirements and ensuring that the language is clear and easy to understand.
[0098] Through the above steps, the generation capability of the large language model is utilized to greatly improve the efficiency of insurance clause generation, and also ensure the quality and market adaptability of the clauses, enabling insurance companies to respond more flexibly to market demands and legal changes.
[0099] The present invention relates to a method for automatically generating insurance clauses, which aims to solve the problem of how to efficiently and accurately generate insurance clauses through automated means. First, by collecting feedback information from insurance users and market research data on insurance projects, these data are deeply analyzed to generate structured insurance data analysis conclusions. These conclusions provide a reliable market basis for the subsequent generation of clauses. Next, the existing insurance clauses and the generated data analysis conclusions are used to form prompt words, which will be used as input, and the large language model will generate target insurance clauses based on the prompt words. The generated clauses cover the scope of protection, exemption clauses and compensation standards, ensuring the comprehensiveness of the clauses and the market adaptability. The present invention improves the design efficiency and compliance of insurance clauses through intelligent means, and provides technical support for the automatic customization of insurance products.
[0100] Furthermore, a second embodiment of the method for automatically generating insurance clauses of the present invention is proposed. In the above step S30, after generating the target insurance clauses, the method further includes:
[0101] Step a1, collecting insurance laws and regulations and insurance industry standards, and converting the insurance laws and regulations and insurance industry standards into logical expressions according to preset rules;
[0102] Step a2, forming insurance rules according to the logical expression, constructing an insurance rule base, and using the insurance rule base to conduct compliance review on the target insurance clause.
[0103] In this embodiment, the latest insurance-related laws and regulations are obtained from national and local legal databases, such as the Insurance Law, the Consumer Rights Protection Law, etc. In addition, the latest guidance and notices from financial regulatory authorities are followed to obtain the latest industry compliance requirements.
[0104] Obtain industry standards from insurance industry associations or professional organizations. These standards may cover product design, risk management, claims payment process, etc. Standards may be obtained through publicly released documents or through communication with industry experts.
[0105] Analyze the collected laws, regulations and industry standards to identify key clauses and requirements. For example, identify the clear requirements for exemption clauses in a legal provision, or the specific provisions on compensation standards in industry standards.
[0106] Convert the parsed clauses into logical expressions that the system can process. For example, "IF the clause contains 'disclaimer', THEN it must contain a specific description and scope of application." These logical expressions can be defined as simple conditional statements or more complex rule combinations.
[0107] The generated logical expressions and rules are stored in a rule base to form a dynamically updated database. The rule base needs to be regularly maintained and updated to ensure that its content is always consistent with the latest laws, regulations and industry standards. Establish a rule management system that allows system administrators to add, modify or delete rules to ensure the accuracy and timeliness of the rule base.
[0108] The generated target insurance clauses are compared with the relevant rules in the rule base for review. The system automatically checks whether there are any content in the clauses that violate legal provisions or whether there are any parts that do not meet industry standards. Based on the review results, a compliance report is generated. If the clauses meet all the rules, the report will mark them as compliant. If there are any problems, the report will list the non-compliant parts in detail and provide corresponding correction suggestions.
[0109] Example description:
[0110] Assume that an insurance company uses the automatic generation method of the present invention to develop a new automobile insurance product. The company first obtains a large amount of user feedback information and market research data, and uses this data to generate structured analysis conclusions. These conclusions show that customers are dissatisfied with the current compensation standards, especially when it comes to third-party liability, and hope to have clearer exemption clauses.
[0111] Specific steps may include:
[0112] Prompt word generation: The system generates prompt words based on user feedback and market data analysis conclusions. For example, the prompt word may be "Generate an automobile insurance policy that contains detailed coverage of third-party liability, clearly defines the compensation standards, and adds exemption clauses for uninsured vehicles."
[0113] Target clause generation: After the prompt word is input into the large language model, the system automatically generates a new automobile insurance clause, which lists in detail the coverage of third-party liability, specific compensation standards, and adds a new exemption clause for uninsured vehicles.
[0114] The system then collects relevant laws and regulations from multiple authoritative sources, such as the Road Traffic Safety Law, the Insurance Law, and industry standards, such as the Insurance Industry Association’s Compensation Standard Guidelines. These regulations and standards contain strict provisions for insurance contract terms, for example, exemption clauses must be clear and not misleading.
[0115] Logical expression conversion: The system converts the provisions on third-party liability insurance in the Road Traffic Safety Law into logical expressions. For example, "IF the coverage includes 'third-party liability', THEN the compensation standard must be clear and must not be lower than the minimum limit prescribed by law."
[0116] Rule base construction: These logical expressions are stored in the rule base to form a rule base that contains the latest laws, regulations and industry standards. This library will be used to conduct compliance review of the generated insurance clauses.
[0117] Clause comparison and review: The system compares and reviews the generated target insurance clauses with the laws and regulations in the rule base. The system finds that there is a content that is not clearly stated in the exemption clause, which may lead to misunderstanding by consumers.
[0118] Compliance report generation: The system generates a compliance report, indicating which clauses are non-compliant and providing specific suggestions for corrections, such as: "Add a definition of 'uninsured vehicle' to the exemption clause and clarify the scope of application of the exemption."
[0119] Amend the terms: Based on the amendment suggestions in the compliance report, the system automatically or manually amends the insurance terms to obtain a new version of the insurance terms.
[0120] After the revised terms pass the compliance review, the system further evaluates the overall risk of the terms and sends the new version of the insurance terms together with the compliance report to the legal and business reviewers. After the reviewers confirm that the terms are correct, the terms can be officially released and applied to the market.
[0121] Through the above steps, this embodiment ensures that the generated insurance clauses not only meet market demand, but also comply with laws, regulations and industry standards, thereby reducing legal risks and improving the credibility and market competitiveness of the clauses.
[0122] Furthermore, a third embodiment of the method for automatically generating insurance clauses of the present invention is proposed, which includes, after the above step a2, the following steps:
[0123] Step b1, generating a compliance report for the target insurance clause. When the content of the compliance report is non-compliant, the compliance report includes suggestions for amending the target insurance clause.
[0124] In this embodiment, after completing the compliance review of the target insurance clauses, the system will generate a compliance report. This report records the results of the review in detail, including which clauses comply with regulations and which clauses do not comply with regulations. If any non-compliance is found in the clauses, the compliance report will include specific correction suggestions. This process not only ensures the legal compliance of the insurance clauses, but also provides guidance for further revision of the clauses.
[0125] A specific implementation method may be: after the target insurance clauses are generated, the system uses the previously constructed insurance rule base to conduct a detailed compliance review of these clauses. The rule base contains logical expressions extracted from laws, regulations and industry standards to check whether the clauses meet all necessary requirements. Through an automated review process, the system compares the target insurance clauses with the rules in the rule base one by one to identify potential non-compliance.
[0126] The system automatically generates a compliance report based on the review results. The report lists which clauses comply with regulations and details the clauses that are non-compliant. For non-compliant parts, the system will clearly indicate which specific laws or industry standards have been violated.
[0127] If certain clauses are marked as non-compliant in the report, the system will generate correction suggestions. Correction suggestions provide specific modification plans for each non-compliant clause based on the best practices and legal requirements in the rule base. Correction suggestions may include redefining certain terms in the clause, adding or deleting specific content, or adjusting the clause structure to ensure compliance.
[0128] The system can automatically modify non-compliant clauses based on the correction suggestions in the compliance report. For example, if the exemption scope of a clause is too broad, the system can add more specific exemption conditions based on the correction suggestions.
[0129] For complex or high-risk amendments, the system can generate a report and submit it to legal experts or auditors for manual review. The auditors will further adjust the terms based on the amendment suggestions to ensure that the final version of the insurance terms is fully compliant.
[0130] After the clauses are revised, the system will review them again using the rule base to ensure that all revised clauses comply with legal requirements and industry standards. This re-review step ensures that the final insurance clauses meet both compliance requirements and market demand.
[0131] Example description:
[0132] A health insurance company launched a new health insurance product that mainly covers hospitalization expenses, outpatient expenses, and routine physical examination expenses. Through user feedback and market research data analysis, the company determined that customers wanted to increase the coverage of routine physical examination expenses and wanted a clearer description of the exemption clause.
[0133] First, a prompt word is generated: "Generate a health insurance clause that covers hospitalization expenses, outpatient expenses, and routine physical examination expenses, and clearly states that physical examinations are not covered." Then, the prompt word is input into the large language model to generate a complete health insurance clause, which details the coverage of hospitalization, outpatient and physical examination expenses, and lists several exemptions.
[0134] The system automatically collected relevant regulations and standards from the national legal database and insurance industry standards, such as the Basic Medical Insurance Provisions and the Consumer Rights Protection Law. These regulations require that insurance clauses must be clear and unambiguous when describing the coverage and exemptions to avoid misleading information. The system converted these regulations into logical expressions and built a compliance rule library.
[0135] The generated health insurance clauses were input into the system, and the system used the rule base to conduct a compliance review. The review results showed that some of the exemption clauses were non-compliant. The specific problem was that the clauses failed to clearly list which physical examination items were not covered, which might cause customers to misunderstand the coverage.
[0136] The system automatically generated a compliance report, which clearly pointed out the shortcomings of the exemption clause and made detailed correction suggestions:
[0137] Correction suggestion 1: In the disclaimer, add a clear list of items that are not covered, such as cosmetic examinations, non-essential health examinations, etc.
[0138] Correction suggestion 2: Use clear legal terms, such as "This protection does not apply to the following medical examinations: cosmetic examinations, fitness examinations, etc." to ensure that customers can understand the actual meaning of the terms.
[0139] Through the above steps, this embodiment ensures that the generated insurance clauses are subject to strict compliance review before being released, thereby reducing legal risks and improving market trust and customer satisfaction.
[0140] Furthermore, a fourth embodiment of the method for automatically generating insurance clauses of the present invention is proposed. In the above step b1, after generating a compliance report for the target insurance clause, the method further includes:
[0141] Step c1, when the content of the compliance report is non-compliant, the target insurance clause is revised according to the revision suggestion to obtain a new version of the insurance clause, and the revision record of the new version of the insurance clause is retained.
[0142] In this embodiment, after generating a compliance report for the target insurance clause, if the report indicates that the clause is non-compliant, the system will revise the target insurance clause according to the revision suggestions in the compliance report. The revised new version of the insurance clause will be saved, and all revision records will be retained. This process ensures that the insurance clause not only complies with the requirements of laws and regulations, but also tracks and manages all revision histories of the clause, providing a transparent modification path and audit trail.
[0143] A specific implementation method may be: the system generates a compliance report based on the results of the compliance review, which lists in detail which clauses have non-compliance issues. For example, some clauses may not meet the latest legal requirements, or some descriptions are too vague and easily lead to ambiguity.
[0144] The system provides detailed correction suggestions in the report, based on the best practices and regulatory requirements in the rule base. These suggestions can be specific changes to the content of the clauses, adding explanatory text, or deleting unnecessary parts.
[0145] The system automatically modifies non-compliant clauses based on the correction suggestions in the compliance report. For example, if the description in the exemption clause is too broad, the system can add specific restrictions based on the correction suggestions to make it comply with legal requirements.
[0146] Manual revision (optional): For complex or high-risk clauses, the system can submit compliance reports and revision suggestions to the legal or compliance team for manual review and modification. The manually revised clauses are re-entered into the system for review to ensure compliance.
[0147] The revised policy will be saved as a new version. This new version should fully meet the compliance requirements and be ready for the next step, such as re-review or release.
[0148] The system records the differences between the revised clauses and the previous versions and generates a revision log. These records include the revision time, revision content, reason for revision, and whether further compliance review has been passed. By retaining all revision records, the system can provide a detailed audit trail to facilitate future review or evaluation of clauses.
[0149] After the new version of the insurance terms is generated, the system will conduct another compliance review to ensure that all amendments meet the requirements. Only the terms that pass the review will be considered the final version and enter the release process.
[0150] All amendment processes and new versions of insurance clauses will be archived to ensure that the amendment history of each clause can be traced back in the future. This not only helps internal management, but also provides support for external audits.
[0151] After passing the compliance review, the revised terms can be submitted to the business department for publication or further promotion.
[0152] Example description:
[0153] When reviewing the newly generated auto insurance clauses, the system found that the exemption clause "no compensation for accidents occurring on private roads" lacked a clear definition and was prone to disputes. Based on the revision suggestions in the compliance report, the system modified the clause and added explanations such as "private roads include but are not limited to private driveways owned by individuals, unopened construction roads, etc." The system generated a revised new version of the clause, retained all revision records, and conducted another compliance review to ensure the compliance and clarity of the clause.
[0154] Through the above steps, this embodiment can efficiently and accurately amend non-compliant insurance clauses and ensure that all amendment processes are transparent and traceable, which helps insurance companies manage the compliance and legal risks of the clauses.
[0155] Furthermore, a fifth embodiment of the method for automatically generating insurance clauses of the present invention is proposed. In the above step c1, after obtaining the new version of the insurance clauses, the method further includes:
[0156] Step d1, evaluating the overall risk level of the target insurance clause, when the overall risk level is greater than a preset threshold, sending the target insurance clause, amendment suggestions and new version of the insurance clause to the reviewer for review.
[0157] In this embodiment, after the target insurance clauses are revised and a new version of the clauses is generated, the system needs to evaluate the overall risk level of these clauses. The overall risk assessment is to determine the potential legal and business risks that these clauses may cause in actual application. If the risk assessment results show that the overall risk level exceeds the preset safety threshold, the system will submit the target insurance clauses, amendment suggestions, and the new version of the insurance clauses to manual reviewers to ensure that the clauses are reviewed more deeply before release.
[0158] Specific implementations may be:
[0159] Risk Assessment Model: The system uses a pre-set risk assessment model to conduct an overall risk assessment of the target insurance terms. The assessment model takes into account a variety of factors, including but not limited to the complexity of the terms, legal compliance, market response, user complaint history, and risk trends in the insurance industry.
[0160] Calculate risk score: The system will calculate an overall risk score through quantitative and qualitative analysis of these factors. The higher the score, the greater the potential risk of the clause. This score is compared with the preset risk threshold.
[0161] Setting thresholds: In the system settings, insurance companies can set a threshold for the overall risk score. If the risk score of a certain clause exceeds this threshold, it indicates that the clause may have greater legal or market risks in practice and requires further manual review.
[0162] Automatic trigger mechanism: When the risk score exceeds the threshold, the system will automatically trigger the next steps and submit the relevant content to the reviewer.
[0163] Material preparation: The system packages the target insurance clauses, amendment suggestions and the new version of the insurance clauses into a complete review material package to ensure that the reviewers can fully understand the modification background, modification content and potential risks of the clauses.
[0164] Send to reviewer: The system submits these materials to the designated reviewer or review team. Reviewers include legal advisors, compliance officers, insurance product managers, etc. They will conduct a comprehensive review of the terms to ensure that the final version of the insurance terms not only meets legal requirements but also effectively controls risks.
[0165] Manual review: Reviewers conduct a detailed review of the submitted materials to confirm the legality, compliance, and market suitability of the terms. If reviewers find further risks or potential problems, they can make new amendment suggestions or request changes to the terms.
[0166] Feedback and Adjustment: Based on the feedback from the reviewers, the system may need to adjust the terms again, and then evaluate and review them again until the terms are confirmed to have no major risks and meet the release conditions.
[0167] Example description:
[0168] An insurance company has launched a new auto insurance product that includes third-party liability insurance, vehicle damage insurance, and additional roadside assistance services. After the terms are generated and the initial compliance review is conducted, the system enters the overall risk assessment stage. Since the insurance terms involve more complex exemption clauses and high-risk scenarios (such as accident exemption on private roads), the company hopes to conduct a detailed risk assessment through an automated system and submit it for manual review when necessary.
[0169] The system first generated the target automobile insurance terms including the following:
[0170] Third-party liability insurance: covers liability compensation for third-party personnel and property losses.
[0171] Vehicle damage insurance: covers compensation for vehicle losses caused by natural disasters or traffic accidents.
[0172] Roadside Assistance Service: Provides roadside assistance service in the event of an accident or vehicle breakdown, but exempts liability for accidents occurring on private roads.
[0173] During the initial compliance review, the system found that these terms basically complied with industry standards and legal requirements, but the description of the exemption clauses for private roads was not specific enough.
[0174] According to the amendments recommended by the preliminary compliance review, the system modified the terms, added a specific definition of "private road" (such as private driveways, unopened construction land, etc.), and refined the wording in the exemption clause. After the amendment, the system generated a new version of the insurance terms and recorded all the changes and reasons.
[0175] The system then initiates an overall risk assessment process to evaluate the potential risks of the new version of the terms. The following are the key steps of the assessment process:
[0176] Evaluation factors:
[0177] Legal risks: Evaluate whether the terms may violate existing laws and regulations, especially those involving third-party liability and disclaimer clauses.
[0178] Market risk: Analyze whether the terms will affect customer acceptance, and whether complex exemption clauses may lead to customer complaints or a decline in market competitiveness.
[0179] Operational risk: Assess whether the execution of the terms in actual operations may lead to legal disputes or compensation problems.
[0180] Risk score calculation:
[0181] Based on the above factors, the system uses quantitative analysis methods to calculate the score of each risk factor and comprehensively obtains the overall risk score of the clause. This score is higher than the safety threshold set by the company, indicating that the clause may have greater legal and market risks in practice.
[0182] Since the risk score exceeds the preset safety threshold, the system automatically packages the target insurance clauses, new version clauses, amendment suggestions and compliance reports into a complete material package and sends it to the legal advisor and product review team for in-depth review.
[0183] Legal advisor review: After receiving the materials, the legal advisor conducted a detailed review of the terms, paying special attention to the definition of "private roads" involved in the exemption clause, confirming its legality and possible legal risks. The legal advisor believes that although the definition has been improved, it may still cause misunderstandings among some customers, and it is recommended to further clarify the applicable scenarios of the exemption clause.
[0184] Product Review Team Assessment: The Product Review Team assessed the competitiveness of the terms in the market and noted that certain terms may affect customer experience, especially in terms of the limited scope of roadside assistance services. The team recommended that the terms of the assistance service be re-evaluated to increase customer satisfaction and market competitiveness.
[0185] Feedback and Amendment: Based on the feedback from the legal advisor and product review team, the system further amended the terms, including clearer disclaimers and adjustments to the rescue service terms. The revised terms were reassessed for overall risk, and the score dropped below the threshold, with risk control within an acceptable range.
[0186] Final confirmation and release: The revised new version of the insurance terms passed all review steps and was confirmed to be in line with legal requirements, market demand and the company's risk management strategy. The terms were finally approved for release and officially applied to the company's new auto insurance products.
[0187] Through the above steps, this embodiment can effectively manage and control the overall risk of insurance clauses, ensuring that the clauses can withstand the test both legally and in the market, thereby reducing potential legal disputes and market complaints.
[0188] Furthermore, a sixth embodiment of the method for automatically generating insurance clauses of the present invention is proposed. In the above step d1, after generating a compliance report for the target insurance clause, the method further includes:
[0189] Step e1, when the content of the compliance report is compliant, the target insurance clause is sent to the business personnel.
[0190] In this embodiment, after the compliance report is generated, if the report shows that the target insurance clauses have complied with all laws, regulations and industry standards, the system will further send these compliant clauses to the business personnel. The business personnel will perform subsequent operations based on these clauses, such as incorporating the clauses into formal insurance products, formulating marketing strategies or communicating with customers.
[0191] Specific implementations may be:
[0192] Compliance check: The system conducts a detailed compliance check on the generated target insurance clauses to ensure that the clauses comply with laws and regulations, industry standards and internal company policies.
[0193] Compliance report: If the content of the clause meets all requirements, the system generates a compliance report marked as "Compliant", detailing the basis for the clause to pass the review and the referenced regulatory provisions.
[0194] Automatic confirmation: After the system generates a compliance report, it automatically checks the status of the report. If the report is marked as "compliant", the system will proceed to the next step.
[0195] Manual review (optional): In some cases, compliance reports may require manual review. For example, clauses involving high-risk products, although marked as compliant, may still require final manual confirmation from the legal team or compliance officer.
[0196] Prepare materials: The system packages the compliant target insurance terms and compliance reports and prepares a material package to be sent to business personnel. The material package may include the final version of the insurance terms, compliance reports, revision history, and relevant market analysis data.
[0197] Automatic distribution: The system automatically sends these materials to designated business personnel or teams based on business needs. Business personnel may include product managers, marketing teams, customer service teams, etc. They will use these terms to advance subsequent business processes.
[0198] Subsequent processing: After receiving compliant insurance terms, business personnel may incorporate them into the formal insurance contract, design corresponding marketing strategies, or develop customer communication plans.
[0199] Business feedback: When business personnel are actually applying these terms, they may find areas that need further optimization. The system can record this feedback as a reference for future generation and adjustment of terms.
[0200] Circular improvement: The system regularly updates the rule base and compliance review process based on feedback from business personnel and actual market performance to ensure that the generated insurance terms always meet the latest market needs and laws and regulations.
[0201] Example description:
[0202] After a certain automobile insurance clause is generated, it passes the compliance check of the system, and the compliance report generated by the system marks the clause content as compliant. The system automatically sends the clause to the product management team.
[0203] After receiving these terms, the product management team integrated them into existing auto insurance contracts and arranged for the customer service team to update customer communication materials to ensure that all newly signed customers understood these terms.
[0204] Through the above steps, this embodiment not only ensures the legality and compliance of insurance clauses, but also helps insurance companies to efficiently apply compliant clauses to actual business, thereby improving operational efficiency and market competitiveness.
[0205] Furthermore, a seventh embodiment of the method for automatically generating insurance clauses of the present invention is proposed, which includes, after the above step e1, the following steps:
[0206] Step f1: Evaluate the target insurance clause using the large language model at preset time intervals, and adjust the content of the target insurance clause according to the evaluation results.
[0207] In this embodiment, after the target insurance clauses are sent to business personnel and begin to be applied, the dynamic nature of the clauses and changes in the market may cause the clauses to need to be adjusted regularly. To ensure that the insurance clauses always meet the latest market needs and legal requirements, the system will use a large language model to periodically evaluate the published target insurance clauses within a preset time period. Based on the evaluation results, the system will make timely adjustments to the content of the clauses to ensure the continued effectiveness and competitiveness of the clauses.
[0208] Specific implementations may be:
[0209] Time interval setting: System administrators or business teams can set the evaluation interval based on the characteristics of insurance products and market environment. For example, some high-frequency changing markets may require monthly evaluation, while more stable products may require quarterly or semi-annual evaluation.
[0210] Automatic trigger mechanism: The system automatically triggers the evaluation process of the target insurance terms according to the set time interval. This mechanism ensures that the terms can be reviewed and updated regularly.
[0211] Clause analysis: At preset time intervals, the system uses a large language model to conduct in-depth analysis of the target insurance clauses. The analysis includes market feedback, new laws and regulations, changes in competitors' clauses, and customer complaints and opinions.
[0212] Identify change needs: The large language model can identify parts of the clauses that may need to be adjusted, such as when certain clauses are controversial in the market or when new legal requirements affect the compliance of the clauses. The model can also recommend enhancing or deleting certain clauses based on market trends.
[0213] Generate correction suggestions: Based on the analysis results of the large language model, the system will generate correction suggestions. These suggestions may include refining the wording of certain clauses, adjusting the compensation standards, adding new coverage, or deleting exemption clauses that are no longer applicable.
[0214] Automatic and manual revision: The system can automatically adjust the terms based on the revision suggestions, and submit the suggestions to the business team or legal experts for manual review when necessary to ensure that the adjusted terms still comply with legal and market requirements.
[0215] Dynamic update: The adjusted terms will be re-issued and re-evaluated at the next preset time period. This dynamic update process ensures that the insurance terms are always in the best state and can respond to changes in the market and laws.
[0216] Feedback mechanism: The system can also further optimize the evaluation model and adjustment mechanism based on customer feedback and market performance, thereby improving the accuracy and timeliness of terms adjustments.
[0217] Example description:
[0218] For an automobile insurance product, the system is set to conduct an assessment once a month. In one assessment, the large language model identified that a competitor’s new terms added special protection for “new energy vehicles”, while the company’s existing terms did not make clear provisions for this. The system generated a correction suggestion, suggesting the addition of special terms for new energy vehicles to maintain market competitiveness. After review, the new terms were released the following month, quickly adapting to market changes.
[0219] Through regular evaluation and dynamic adjustment, the system of this embodiment can ensure that the insurance terms not only comply with current legal requirements, but also actively adapt to market changes and enhance the long-term competitiveness of the product and customer satisfaction.
[0220] The present invention also provides a device for automatically generating insurance clauses.
[0221] Reference Figure 3 , Figure 3 This is a functional module diagram of the first embodiment of the device for automatically generating insurance clauses of the present invention. The device for automatically generating insurance clauses includes:
[0222] The data collection and analysis module obtains feedback information from insurance users and market research data of insurance projects and performs data analysis to generate structured insurance data analysis conclusions;
[0223] A prompt word generation module obtains existing insurance terms and forms prompt words according to the existing insurance terms and the insurance data analysis conclusion;
[0224] The insurance clause generation module inputs the prompt words into the large language model to generate target insurance clauses, wherein the content of the target insurance clauses includes the coverage, exemption clauses and compensation standards.
[0225] In addition, an embodiment of the present invention further provides a computer storage medium.
[0226] The storage medium stores an insurance clause automatic generation program, and when the insurance clause automatic generation program is executed by the processor, the steps of the insurance clause automatic generation method described above are implemented.
[0227] The specific implementation of the storage medium of the present invention is basically the same as the above-mentioned embodiments of the method for automatically generating insurance terms, and will not be repeated here.
[0228] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.
[0229] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.
[0230] In addition, each functional module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. If the function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, RandomAccess Memory), a disk or an optical disk.
[0231] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0232] The above are only various implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for automatically generating insurance clauses, characterized in that: The following steps are involved: Obtain feedback information from insurance users and insurance project market research data and conduct data analysis to generate structured insurance data analysis conclusions; Obtaining existing insurance terms, and forming prompt words according to the existing insurance terms and the insurance data analysis conclusion; The prompt words are input into a large language model to generate target insurance clauses, wherein the contents of the target insurance clauses include coverage, exemption clauses and compensation standards.
2. The method for automatically generating insurance clauses according to claim 1, characterized in that: After generating the target insurance terms, it also includes: Collect insurance laws and regulations and insurance industry standards, and convert the insurance laws and regulations and insurance industry standards into logical expressions according to preset rules; Insurance rules are formed according to the logical expressions, an insurance rule base is constructed, and the insurance rule base is used to conduct compliance review on the target insurance clauses.
3. The method for automatically generating insurance clauses according to claim 2, characterized in that: After the target insurance clause is reviewed for compliance using the insurance rule base, the following steps are further included: A compliance report is generated for the target insurance clause. When the content of the compliance report is non-compliant, the compliance report includes a suggestion for amending the target insurance clause.
4. The method for automatically generating insurance clauses according to claim 3, characterized in that: After generating a compliance report for the target insurance clause, it also includes: When the content of the compliance report is non-compliant, the target insurance clause is amended according to the amendment suggestion to obtain a new version of the insurance clause, and the amendment record of the new version of the insurance clause is retained.
5. The method for automatically generating insurance clauses according to claim 4, characterized in that: After obtaining the new version of the insurance terms, it also includes: The overall risk level of the target insurance clause is evaluated. When the overall risk level is greater than a preset threshold, the target insurance clause, amendment suggestions and the new version of the insurance clause are sent to the reviewer for review.
6. The method for automatically generating insurance clauses according to claim 3, characterized in that: After generating a compliance report for the target insurance clause, it also includes: When the content of the compliance report is compliant, the target insurance clause is sent to the business personnel.
7. The method for automatically generating insurance clauses as claimed in claim 6, characterized in that: After sending the target insurance clause to the business personnel, it also includes: At each preset time period, the target insurance clause is evaluated using the large language model, and the content of the target insurance clause is adjusted according to the evaluation result.
8. An automatic insurance clause generation device, characterized in that: The insurance clause automatic generation device comprises: The data collection and analysis module obtains feedback information from insurance users and market research data of insurance projects and performs data analysis to generate structured insurance data analysis conclusions; A prompt word generation module obtains existing insurance terms and forms prompt words according to the existing insurance terms and the insurance data analysis conclusion; The insurance clause generation module inputs the prompt words into the large language model to generate target insurance clauses, wherein the content of the target insurance clauses includes the coverage, exemption clauses and compensation standards.
9. An automatic insurance clause generation device, characterized in that: The insurance clause automatic generation device includes a memory, a processor, and an insurance clause automatic generation program stored in the memory and executable on the processor. When the insurance clause automatic generation program is executed by the processor, the steps of the insurance clause automatic generation method as described in any one of claims 1 to 7 are implemented.
10. A computer storage medium, characterized in that: The storage medium stores an insurance clause automatic generation program, and when the insurance clause automatic generation program is executed by the processor, the steps of the insurance clause automatic generation method according to any one of claims 1 to 7 are implemented.
