Insurance product recommendation processing method and device, computer equipment and medium

Abstract

The invention relates to the technical field of artificial intelligence, discloses an insurance product recommendation processing method and device based on artificial intelligence, equipment and a medium, and aims to solve the problem of low effectiveness of technical means of insurance product recommendation in the traditional technology. According to a'personal innate attribute-traditional medicine-modern medicine 'ternary mapping relation, health risks corresponding to personal innate attributes are predicted on the basis of a retrieval module of a preset retrieval enhancement generation scheme, then corresponding insurance products are matched, and finally, the health risks corresponding to the personal innate attributes are generated on the basis of corresponding generation modules contained in the preset retrieval enhancement generation scheme. According to the method, the target insurance product scheme corresponding to the preset insurance individual customer is generated, so that the credibility, personalization, accuracy, timeliness and practicability of insurance product recommendation can be improved, the increasing effective recommendation requirements corresponding to personalized insurance products are met, and the effectiveness of the technical means of insurance product recommendation is improved.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence technology and financial technology, and in particular to an artificial intelligence-based insurance product recommendation processing method, apparatus, equipment and medium. Background Technology

[0002] Insurance recommendation services aim to assess user needs and risks based on data analysis, matching users with suitable insurance products to help them effectively transfer potential financial losses through personalized insurance plans. Traditional technologies typically match users with suitable insurance products based on structured data such as age, income, occupation, medical history, and health checkup information, using rule engines to perform a simple mapping between diseases and insurance products.

[0003] The inventors realized that in traditional technologies, matching insurance products for users mainly relies on existing explicit structured data and rule engine recommendation methods. However, due to the limitations and lag of information in the explicit structured data, and the mechanical matching based on rule engine recommendation methods, it is impossible to comprehensively and accurately reflect the dynamic diversity of users' health risk factors and characteristics. It also leads to a lag in responding to the potential diversity of users' health risk factors and characteristics. As a result, the insurance product recommendations lack personalization, accuracy, and practicality, making it difficult to meet the growing demand for effective recommendations of personalized insurance products and reducing the effectiveness of the aforementioned technical means of insurance product recommendation.

[0004] Therefore, how to improve the effectiveness of technical means for recommending insurance products has become an urgent technical problem to be solved in the insurance industry. Summary of the Invention

[0005] This invention provides an insurance product recommendation processing method, apparatus, computer equipment, and medium based on artificial intelligence, to solve the technical problem of low effectiveness of traditional insurance product recommendation techniques.

[0006] Firstly, an artificial intelligence-based insurance product recommendation processing method is provided, comprising: determining the innate endowment attributes corresponding to a preset insurance individual customer; predicting the health risk corresponding to the innate endowment attributes based on the innate endowment attributes and a preset Chinese Classics-Modern Medicine information database, and based on a preset health risk retrieval submodule included in a preset retrieval enhancement generation scheme, to obtain the estimated health risk corresponding to the preset insurance individual customer; matching the estimated health risk with a preset insurance product knowledge graph, and based on a preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, to obtain an initial insurance product; embedding the initial insurance product into a preset prompt word template to obtain a target prompt word; generating a target insurance product scheme corresponding to the preset insurance individual customer based on the target prompt word and a preset insurance product scheme generation module included in the preset retrieval enhancement generation scheme; and pushing the target insurance product scheme to the preset insurance individual customer.

[0007] Secondly, an artificial intelligence-based insurance product recommendation processing device is provided, comprising: a first determining module, used to determine the innate endowment attributes corresponding to a preset insurance individual customer; a health risk prediction module, used to predict the health risk corresponding to the innate endowment attributes based on the innate endowment attributes and a preset Chinese Classics-Modern Medicine information database, and based on a preset health risk retrieval submodule included in a preset retrieval enhancement generation scheme, to obtain the estimated health risk corresponding to the preset insurance individual customer; an insurance product matching module, used to match the estimated health risk with the preset insurance product knowledge graph, and based on a preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, to obtain an initial insurance product; a prompt word embedding module, used to embed the initial insurance product into a preset prompt word template to obtain a target prompt word; an insurance plan generation module, used to generate a target insurance product plan corresponding to the preset insurance individual customer based on the target prompt word and a preset insurance product plan generation module included in the preset retrieval enhancement generation scheme; and an insurance plan push module, used to push the target insurance product plan to the preset insurance individual customer.

[0008] 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.

[0009] 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.

[0010] The aforementioned insurance product recommendation processing method, device, computer equipment, and storage medium implement a solution whereby the method determines the innate attributes of a pre-defined insurance customer, predicts the health risks corresponding to the innate attributes based on the three-element mapping relationship of "innate attributes - traditional medicine - modern medicine," and uses a pre-defined retrieval enhancement generation scheme. It then matches the corresponding insurance products and finally generates a target insurance product plan for the pre-defined insurance customer. This not only combines innate attributes, traditional Chinese medicine, and other aspects of health knowledge with modern medicine, but also constructs a multi-source heterogeneous knowledge enhancement generation system based on a retrieval enhancement generation architecture. This system integrates traditional Chinese health knowledge, modern medicine, insurance product knowledge, and the user's birth date, achieving insurance product recommendations based on the correlation between "innate attributes - disease risk - insurance" and the fusion of multi-source heterogeneous information data. This improves the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations, enhancing the effectiveness of the aforementioned technical means of insurance product recommendation. Attached Figure Description

[0011] 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.

[0012] Figure 1 A flowchart illustrating the artificial intelligence-based insurance product recommendation processing method provided in an embodiment of the present invention;

[0013] Figure 2 This is a schematic diagram of the first sub-process of the insurance product recommendation processing method based on artificial intelligence provided in an embodiment of the present invention;

[0014] Figure 3 A schematic diagram of the second sub-process of the AI-based insurance product recommendation processing method provided in an embodiment of the present invention;

[0015] Figure 4 A schematic block diagram of an artificial intelligence-based insurance product recommendation processing device provided in an embodiment of the present invention;

[0016] Figure 5 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention;

[0017] Figure 6 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0018] 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.

[0019] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0020] This invention provides an artificial intelligence-based insurance product recommendation processing method. The method can be applied to computer devices including but not limited to smartphones, tablets, desktop computers, servers, cloud platforms, etc., and can be used in the fields of finance and insurance to recommend insurance products based on artificial intelligence.

[0021] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0022] Please see Figure 1 , Figure 1 This is a flowchart illustrating an artificial intelligence-based insurance product recommendation processing method provided in an embodiment of the present invention. Figure 1 As shown, the method includes, but is not limited to, the following steps S11-S16:

[0023] S11. Determine the innate attributes of the individuals corresponding to the pre-set insurance customers.

[0024] Interpretatively, this involves determining the innate endowment attributes of a pre-defined insurance client. These attributes represent the sum of the client's inherent physiology and developmental potential, determined by factors such as genetics, the birth environment, and embryonic development conditions, forming the fundamental basis of an individual's life development. Innate endowment attributes include, but are not limited to, Traditional Chinese Medicine (TCM) constitution, TCM Five Elements constitution, and Five Elements constitution. TCM constitution and Five Elements constitution can be determined by referencing relevant knowledge from traditional Chinese classics and medicine, including but not limited to the *Huangdi Neijing* and *Nanjing*. These classics and medical texts have been passed down and developed for at least thousands of years, possessing a certain degree of practical accuracy and reliability; otherwise, they would have long been eliminated by nature. Therefore, TCM constitution can be determined by professional traditional medical knowledge and personnel, while Five Elements constitution can be determined based on traditional Chinese classics. For example, the Five Elements constitution can be determined in the following way: determine the birth date corresponding to the preset insurance individual customer; determine the customer's birth date and time information data according to the birth date; convert the customer's birth date and time information data into Yin-Yang and Five Elements attributes based on the preset birth date and time-Five Elements conversion model to obtain the Five Elements weight model; determine the corresponding Five Elements constitution according to the Five Elements weight model.

[0025] Among them, the birth chart (also known as the Four Pillars of Destiny) is a concept in traditional Chinese studies. It is based on the year, month, day, and hour of a person's birth, with each hour corresponding to two characters in the Heavenly Stems and Earthly Branches, totaling eight characters, hence the name "Eight Characters". Therefore, the birth time of a pre-defined insurance client is generally determined, i.e., the birth date, and converted into the Eight Characters to obtain the client's Eight Characters information data. For example, May 15, 1990 at 10:00 AM (the 21st day of the fourth lunar month, the Si hour) is converted into the following Eight Characters: Geng Wu, Xin Si, Geng Chen, Xin Si, i.e., the year pillar is Geng Wu, the month pillar is Xin Si, the day pillar is Geng Chen, and the hour pillar is Xin Si. The Eight Characters conversion involved can refer to existing Eight Characters conversion methods, which will not be elaborated here.

[0026] Furthermore, a pre-set Bazi-Five Elements conversion model is established, which represents the conversion of the birth date and time into the corresponding Yin-Yang and Five Elements. The pre-set Bazi-Five Elements conversion model mainly describes the conversion process of the birth date and time into the corresponding Yin-Yang and Five Elements. The conversion process of the birth date and time into the corresponding Yin-Yang and Five Elements can refer to existing corresponding conversion methods, which will not be elaborated here. Thus, the Five Elements constitution corresponding to the innate endowment attributes of the pre-set insurance individual customer is obtained.

[0027] It should be noted that the above-mentioned determination of the Five Elements constitution is only a way or channel to quantify and express an individual's innate endowment attributes. It is not intended to limit an individual's innate endowment attributes. In other feasible situations, other methods can also be used to quantify and determine the quantification and expression of an individual's innate endowment attributes, such as methods that have been verified to be feasible in practice, including but not limited to pulse diagnosis and facial diagnosis. Furthermore, it should be noted that the determination of the Five Elements constitution, as a processing technique for insurance product recommendations, describes an feasible method for processing insurance product recommendations. Those skilled in the art, based on their basic research and development capabilities and technical understanding, can understand that, as a technical means, its verification is not limited to corresponding training samples, nor should it be limited to limited experiments or trials under modern scientific thinking. More importantly, it should be verified in rich business scenarios of corresponding practice. The Five Elements constitution can be determined based on the aforementioned birth date and time, Five Elements transformation, traditional medicine, and other traditional Chinese cultural content. It also draws on the fact that these traditional Chinese cultural and medical practices have been passed down, verified, and developed for at least thousands of years, possessing a certain degree of practical accuracy and reliability. Its practicality does not apply to determination using modern scientific thinking. Therefore, as an extended approach and technical means to further enrich the processing methods for insurance product recommendations, it possesses practicality and feasibility.

[0028] Based on the above technical concept, setup, and description, according to the client's birth date and corresponding Bazi (Four Pillars of Destiny) information data, and based on a preset Bazi-Five Elements conversion model, the client's Bazi information data is converted into Yin-Yang and Five Elements, resulting in a Five Elements weight model. This determines the corresponding individual innate endowment attributes. In this case, the individual's innate endowment attributes represent the distribution ratio of the Five Elements in the birth time-space environment corresponding to the Bazi, quantifying the strength and balance of the Five Elements corresponding to the individual's innate endowment attributes. This allows for the assessment of the imbalance of the innate endowment attributes of the preset insurance client from the perspective of Five Elements energy. Furthermore, based on the balance and imbalance of the innate endowment attributes corresponding to Five Elements energy, the health risks of the preset insurance client can be predicted. This allows for the quantification of the preset insurance client's constitution and the prediction of their health risks based on traditional Chinese culture and medicine.

[0029] S12. Based on the individual's innate endowment attributes and the preset Chinese Classics-Modern Medicine Information Database, and based on the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, predict the health risk corresponding to the individual's innate endowment attributes, and obtain the estimated health risk corresponding to the preset insurance individual customer.

[0030] Explain, a pre-set information database is established to map the relationship between traditional Chinese medicine and modern medicine. This pre-set database represents the information database of the three-element mapping relationship of "individual innate endowment attributes - traditional medicine - modern medicine". For example, an example of the three-element mapping relationship of "individual innate endowment attributes - traditional Chinese medicine - modern medicine" can be represented as: "Increased earth element in individual innate endowment attributes → phlegm-dampness constitution → elevated triglycerides".

[0031] A pre-set retrieval-augmented generation scheme, also known as a preset retrieval-augmented generation scheme, is a hybrid technology in Natural Language Processing (NLP) that combines information retrieval and generative models. This scheme enhances the accuracy of the generative model's responses by retrieving external knowledge bases, thus addressing the "illusion" (fabricated information) problem inherent in pure generative models. Based on this, this invention combines a preset Chinese Classics-Modern Medicine information database with the retrieval-augmented generation scheme to recommend insurance products. Furthermore, based on the above description and technical concept, the preset retrieval-augmented generation scheme includes a preset health risk retrieval submodule, a preset insurance product retrieval submodule, and a preset insurance product scheme generation module. The preset health risk retrieval submodule represents a scheme based on the aforementioned "individual innate..." The module retrieves information on modern medical health risks, such as diseases, from a pre-defined database of traditional Chinese culture and modern medicine, based on the "endowment attribute - traditional medicine - modern medicine" ternary mapping relationship. The pre-defined insurance product retrieval sub-module retrieves insurance products corresponding to modern medical health risks, such as diseases, from the corresponding information database based on the "modern medicine - insurance product" binary mapping relationship. The pre-defined insurance product scheme generation module generates insurance product schemes based on a generation model of a pre-defined retrieval enhancement generation scheme. The generation model uses models including but not limited to the Large Language Model (LLM) corresponding to GPT, Claude, and PaLM.

[0032] Based on the above concept and setup, and according to an individual's innate endowment attributes and a preset Chinese Classics-Modern Medicine information database, and using the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the individual's innate endowment attributes are used as search conditions to search the preset Chinese Classics-Modern Medicine information database to match the health risks corresponding to the individual's innate endowment attributes. This predicts the health risks corresponding to the individual's innate endowment attributes, resulting in the estimated health risks for the preset insurance individual client. The estimated health risk represents the probability of disease risk in modern medicine based on the predicted physiological constitution of the preset insurance individual client according to their innate endowment attributes. For example, if an individual's innate endowment attributes indicate an enhanced Earth element, the corresponding modern medical health risks will be mapped through the path: "Enhanced Earth Element in Innate Endowment Attribute → Phlegm-Dampness Constitution → Elevated Triglycerides → Cardiovascular and Cerebrovascular Diseases".

[0033] S13. Based on the knowledge graph of the estimated health risk and the preset insurance product, and based on the preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, match the insurance product corresponding to the estimated health risk to obtain the initial insurance product.

[0034] Explained, a pre-set insurance product knowledge graph is used. This pre-set knowledge graph represents a knowledge graph built based on insurance products. It can automatically parse insurance terms through natural language processing (NLP) to extract key insurance product information such as disease coverage, exclusions, and claim conditions. This allows for the determination of the corresponding mapping relationship between modern health risks such as modern diseases and insurance products, thus constructing an insurance product knowledge graph. This provides structured information data support for subsequent insurance product retrieval and personalized insurance product recommendations.

[0035] Based on the above concept and setup, according to the knowledge graph of estimated health risks and preset insurance products, and based on the preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, the insurance products corresponding to the estimated health risks are matched to obtain the initial insurance products. That is, the estimated health risks are used as retrieval conditions to perform corresponding searches in the preset insurance product knowledge graph to obtain the insurance products corresponding to the estimated health risks. This realizes the matching between estimated health risks and insurance products, and obtains the initial insurance products. The initial insurance products represent insurance products that cover the estimated health risks.

[0036] S14. Embed the initial insurance product into a preset prompt word template to obtain the target prompt word.

[0037] Explained, a pre-set prompt template, i.e. a preset prompt template, refers to a template based on a prompt, where the prompt is a piece of text used to guide the aforementioned preset insurance product plan generation module to generate an insurance product plan.

[0038] Based on the above concept and settings, a prompt word template is pre-set. The preset prompt word template includes the template format corresponding to the initial insurance product. Then, the initial insurance product is embedded into the preset prompt word template to obtain the target prompt word.

[0039] S15. Based on the target prompt words and the preset insurance product plan generation module included in the preset search enhancement generation scheme, generate the target insurance product plan corresponding to the preset insurance individual customer.

[0040] Explainedly, based on target prompts and the preset insurance product plan generation module included in the preset search enhancement generation scheme, a target insurance product plan corresponding to a preset individual insurance customer is generated. In this way, the powerful language understanding and generation capabilities of the preset insurance product plan generation module based on a large language model are used to generate the target insurance product plan.

[0041] S16. Push the target insurance product plan to the preset insurance individual customer.

[0042] Explanatoryly, the generated target insurance product plan is pushed to the preset individual insurance customers through relevant APP pages, web pages or mini-program pages, etc. The preset individual insurance customers can then receive the insurance product plan generated based on their birth date and time, realizing the recommendation of insurance products based on the preset individual insurance customer's birth date and time.

[0043] This invention, in its embodiments, determines the innate attributes of a pre-defined insurance client, and based on a ternary mapping relationship of "innate attributes - traditional medicine - modern medicine," and a pre-defined retrieval-enhanced generation scheme, predicts the health risks corresponding to these attributes. It then matches appropriate insurance products and finally generates a target insurance product scheme for the pre-defined insurance client. This method not only combines innate attributes, traditional Chinese medicine, and other aspects of health principles with modern medicine through a ternary mapping and cross-validation of "innate attributes - traditional medicine - modern medicine," but also achieves advanced prediction of potential health risks for the pre-defined insurance client based on traditional Chinese health principles. Compared to the information limitations and lag of traditional technologies based on explicit structured data, this method improves the predictive power, dynamism, and accuracy of the predictions. Furthermore, it constructs a multi-source heterogeneous knowledge enhancement system based on a retrieval-enhanced generation architecture. The system integrates traditional Chinese health knowledge, modern medicine, insurance product knowledge, and users' birth dates and times to generate insurance product recommendations based on the correlation between individual innate attributes, disease risks, and insurance, as well as the fusion of multi-source heterogeneous information data. This enhances the diversity and richness of insurance product recommendations in response to individual user differences and complex risk relationships. Compared to traditional linear mapping recommendation methods based on rule engines, it improves the accuracy, personalization, foresight, and practicality of insurance product recommendations. It helps individual insurance customers effectively transfer potential loss risks (such as health risks and property risks). Based on this, the system improves the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations, meeting the growing demand for personalized insurance products and enhancing the effectiveness of the aforementioned technical means of insurance product recommendation.

[0044] Please see Figure 2 , Figure 2 This is a schematic diagram of the first sub-process of the AI-based insurance product recommendation method provided in an embodiment of the present invention. Figure 2 As shown, in this embodiment, determining the innate personal endowment attributes corresponding to a preset insurance individual customer includes:

[0045] S21. Determine the birth date corresponding to the preset insurance individual customer, and determine the original innate endowment attribute corresponding to the preset insurance individual customer based on the birth date;

[0046] S22. Determine the seasonal information data and regional climate information data corresponding to the residence of the preset insurance individual customer;

[0047] S23. Determine the solar term risk correction factor corresponding to the solar term information data, and determine the regional climate correction factor corresponding to the regional climate information data;

[0048] S24. Based on the solar term risk correction factor and the regional climate correction factor, the original individual innate endowment attribute is corrected to obtain the individual innate endowment attribute corresponding to the preset insurance individual customer.

[0049] Interpretatively, the birth date corresponding to the pre-set insurance individual customer is determined, and as mentioned above, the original innate endowment attribute corresponding to the pre-set insurance individual customer is determined based on the birth date. Thus, the original innate endowment attribute is quantitatively expressed. The original innate endowment attribute represents the sum of the pre-set insurance individual customer's innate physiology and development potential, which is jointly determined by factors such as genetic genes, birth time and space environment, and embryonic development environment conditions.

[0050] The system determines the seasonal and regional climate information data corresponding to the residence of the pre-defined insurance individual customer. This includes seasonal and regional climate information from the customer's birth location or current long-term residence. It also determines the corresponding seasonal risk correction factor for the seasonal information data; different seasonal information data generally correspond to different seasonal risk correction factors. The seasonal risk correction factor represents the adjustment amount by which acquired seasonal factors affect the original individual's innate attributes. Seasonal correction focuses on dynamically adjusting the impact of seasonal factors on the original individual's innate attributes based on the climatic patterns of the seasons, from a time perspective. Furthermore, it determines the corresponding regional climate correction factor for the regional climate information data; different regional climate information data generally correspond to different regional climate correction factors. The regional climate correction factor represents the adjustment amount by which acquired regional and climatic factors affect the original individual's innate attributes. Regional correction focuses on localizing the aforementioned original individual's innate attributes based on regional climate differences, from a spatial perspective.

[0051] Among them, the seasonal risk correction factor and the regional climate correction factor can be manually set by relevant personnel based on the degree of influence of different seasonal factors or regional and climatic factors on the original individual's innate endowment attributes. Alternatively, they can be automatically learned through the corresponding training samples corresponding to the influence of seasonal factors or regional and climatic factors on the original individual's innate endowment attributes. For example, for the regional climate correction factor, convolutional neural networks (CNNs) can be used to extract features from regional climate data (such as temperature, humidity, and rainfall) and establish a dynamic map of climate and original individual innate endowment attributes to identify the risk of imbalance of individual innate endowment attributes under specific climatic conditions. For example, "humid and rainy areas → excessive water vapor → increased risk of kidney diseases".

[0052] Furthermore, for the seasonal risk correction factor, the individual's innate endowment attributes can be dynamically adjusted based on the seasonal period at the time of birth of the insured individual, or the influence of the seasonal period corresponding to the Five Elements and Six Qi theory. For example, the wood attribute is enhanced for those born at the Spring Equinox; for the Jueyin Wind Wood corresponding to the Five Elements and Six Qi, corresponding to the period from the Great Cold to the Spring Equinox, the wood attribute of the individual's innate endowment increases by 15%, and the earth attribute decreases by 5%; during the Great Heat solar term, the fire attribute of the individual's innate endowment increases by 15%. Introducing the seasonal risk correction factor into disease risk assessment and prediction can improve the seasonal adaptability of insurance product recommendations based on the time of birth. For the regional climate correction factor, the regional climate correction factor can be introduced in combination with the climate characteristics of the user's birthplace. For example, "birth at a time or long-term residence in a humid area → enhanced water attribute → increased risk of kidney-related diseases."

[0053] Then, based on the solar term risk correction factor and the regional climate correction factor, the individual's innate endowment attributes are corrected to obtain the innate endowment attributes corresponding to the preset insurance individual customers. Thus, by combining the birth time, solar term factor (annual cycle), and regional factor (regional time and space and environmental input), the innate endowment attributes of individuals can more accurately express the innate tendencies and the influence of the solar term and regional climate on the innate endowment attributes of the preset insurance individual customers.

[0054] This invention, through modifications to the initial innate attributes corresponding to the birth time of a pre-defined insurance client based on seasonal and geographical factors, constructs a multi-dimensional quantitative model encompassing "birth time—seasonal—regional climate—innate attributes—disease risk." This enhances the personalized dynamic adaptability of innate attributes based on seasons and geography, improving the accuracy of the innate attributes of individual pre-defined insurance clients. Furthermore, it enables flexible and accurate individual assessments of health risks for these clients, improving the accuracy of health and disease risk predictions. This, in turn, facilitates insurance product recommendations. This not only transforms traditional "static recommendations" to the "dynamic, culturally-aware, and risk-predictive" recommendations of this invention but also improves the timeliness, accuracy, practicality, and personalization of insurance product recommendations, thereby enhancing the effectiveness of the aforementioned technical means of insurance product recommendation.

[0055] In one embodiment, based on the individual's innate endowment attributes and a preset Chinese Classics-Modern Medicine information database, and using a preset health risk retrieval submodule included in a preset retrieval enhancement generation scheme, the health risk corresponding to the individual's innate endowment attributes is predicted to obtain the estimated health risk corresponding to the preset insurance individual customer, including:

[0056] Based on the individual's innate endowment attributes, determine the individual's innate endowment attribute constitution type corresponding to the preset insurance individual customer;

[0057] Based on the individual's innate physical constitution and attributes, the risk target organs corresponding to the preset insurance individual customer are determined;

[0058] Based on the target organs and the preset mapping path of "personal innate endowment attributes - traditional medicine - modern medicine", and based on the preset Chinese classics - modern medicine information database and the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the modern medical health risk corresponding to the target organs is predicted, and the estimated health risk corresponding to the preset insurance individual customer is obtained.

[0059] Explainingly, a pre-defined mapping path of "individual innate endowment attributes - traditional medicine - modern medicine" is established. This pre-defined mapping path represents the mapping path from traditional individual innate endowment attributes to modern medicine. Traditional medicine serves as an intermediate bridge between individual innate endowment attributes and modern medicine. There is a mapping relationship between individual innate endowment attributes and traditional medicine based on the concept of health in traditional Chinese culture. There is a mapping relationship between traditional medicine and modern medicine based on human body manifestations. Thus, the pre-defined mapping path of "individual innate endowment attributes - traditional medicine - modern medicine" is determined. Here, "human body manifestations" is a general term for the external manifestations of the human body, including when a person is in a disease state. It includes both the state described subjectively by relevant personnel and the signs objectively observed by doctors.

[0060] Based on the above concept and setup, the innate physical constitution type corresponding to the pre-set insurance individual customer is determined according to the individual's innate endowment attributes. The innate physical constitution type represents the constitution type based on the individual's innate endowment attributes. Examples of innate physical constitution types are shown in Table 1 below:

[0061] Table 1

[0062] Personal innate endowment attributes and physique Characteristics of the Eight Characters corresponding to the time of birth Human body correspondence Wood-shaped people The Day Master is Jia / Yi, with Wood being strong. liver, gallbladder Fire-shaped person The Day Master is Bing / Ding, with strong Fire element. Heart, small intestine Earth-type people Sun Lord Wu / Ji, earth is weak Spleen and Stomach Metal type The Day Master is Geng / Xin, with strong Metal element. lungs, large intestine Water-shaped people The Day Master is Ren / Gui, and the Water element is weak. Kidneys and bladder

[0063] Then, based on an individual's innate constitution and physical attributes, the target organs for risk are determined for the pre-set insurance client, as shown in Table 1 above. For example, if the pre-set insurance client is determined to be a Wood-type person, the target organs for risk are liver and gallbladder diseases and health risks related to emotional disorders, and so on. Then, based on the target organs and the preset mapping path of "personal innate endowment attributes - traditional medicine - modern medicine", and based on the preset Chinese classics - modern medicine information database and the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the modern medical health risk corresponding to the target organs is predicted, and the estimated health risk corresponding to the preset insurance individual customer is obtained. That is, the target organs and the preset mapping path of "personal innate endowment attributes - traditional medicine - modern medicine" are used as retrieval conditions to retrieve modern medical health risks from the preset Chinese classics - modern medicine information database, and obtain the estimated health risk corresponding to the preset insurance individual customer. Thus, the three-element mapping path of "personal innate endowment attributes - traditional medicine - modern medicine" is realized. Since insurance products are generally associated with modern medicine, the personal innate endowment attributes are then associated and mapped with insurance products through modern medicine, realizing the technical path of recommending insurance products based on personal innate endowment attributes - traditional medicine and modern medicine - insurance products.

[0064] Furthermore, based on the individual's innate physical constitution type, the risk target organs corresponding to the preset insurance individual customer are determined, including:

[0065] Determine the TCM constitution type corresponding to the individual's innate endowment attribute constitution type;

[0066] Based on the TCM constitution type, the target organs for risk are determined for the pre-set insurance individual customer.

[0067] Specifically, when an individual's innate constitution does not correspond to a specific TCM constitution type, the TCM constitution type corresponding to that innate constitution is determined. Based on this TCM constitution type, the target organs for risk are identified for the pre-defined insurance client. Since TCM constitutions are more closely related to the symptoms in traditional medicine, a more accurate and comprehensive mapping relationship can be established with modern medicine through TCM constitution types and their corresponding clinical manifestations. This allows for comprehensive coverage of health risks corresponding to an individual's innate constitution by modern medicine-based insurance products, improving the accuracy, universality, and personalization of insurance product recommendations. See Table 2 below for details.

[0068] Table 2

[0069] Personal innate endowment attributes and physique Traditional Chinese Medicine constitution Typical characteristics Human body correspondence Wood-type person (excessive wood element) Liver Qi Stagnation Constitution Irritability, hypochondriac pain, etc. liver, gallbladder Fire type (excessive fire element) Yin deficiency constitution Dry mouth, insomnia, flushed face, etc. Heart, small intestine Earth-type person (weak Earth element) Spleen deficiency constitution Indigestion, muscle relaxation Spleen and Stomach Metal type person (excessive metal element) Lung dryness constitution Dry cough, dry skin, etc. lungs, large intestine Water-type person (weak water element) Kidney deficiency constitution Lower back and knee pain, aversion to cold, etc. Kidneys and bladder

[0070] This invention, through mapping and linking the innate attributes of a pre-defined insurance client with traditional and modern medical concepts such as constitution and internal organs, enables the prediction of health risks from innate attributes to modern medical understanding. Based on these modern medical health risks, insurance products are recommended, ultimately establishing a connection between the user's innate attributes and insurance products and providing corresponding recommendations. This allows for proactive prediction of potential health risks for pre-defined insurance clients based on traditional Chinese medicine principles. Compared to the limitations and lag of information based on explicit structured data in traditional technologies, this invention improves the predictive power, dynamism, and accuracy of the predictions. It also enhances the diversity and richness of insurance product recommendations in addressing individual differences and complex risk relationships, improving the accuracy, personalization, foresight, and practicality of insurance product recommendations. Furthermore, it enhances the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations, meeting the growing demand for personalized insurance products and improving the effectiveness of the aforementioned technical means of insurance product recommendation.

[0071] In one embodiment, determining the target organs for risk corresponding to the preset insurance individual customer based on the TCM constitution type includes:

[0072] Determine the preset multimodal health information data corresponding to the preset insurance individual customer;

[0073] Based on the individual's innate constitution type, the TCM constitution type, and the preset multimodal health information data, a health risk profile is created for the preset insurance individual customer to obtain the individual customer health risk profile corresponding to the preset insurance individual customer.

[0074] Based on the individual customer's health risk profile, the risk target organs corresponding to the preset insurance individual customer are determined.

[0075] Interpretatively, the pre-defined multimodal health information data corresponding to the pre-defined insurance individual customer is determined. This pre-defined multimodal health information data includes, but is not limited to, long-term stable biomedical indicators such as physical examination reports (biochemical indicators, imaging results), genetic testing (genetic risk, metabolic characteristics), and real-time continuous physiological signals recorded by wearable devices (heart rate variability, sleep cycles, steps, blood oxygen, body temperature, etc.). The aforementioned multi-source information data undergoes standardization processing and feature fusion. Then, based on the individual's innate constitution type, traditional Chinese medicine constitution type, and the pre-defined multimodal health information data, the pre-defined insurance... The system creates health risk profiles for individual clients, resulting in a comprehensive and multi-dimensional profile of each client's health, which combines traditional Chinese medicine and modern medicine. Based on these profiles, the system identifies the target organs for risk. This approach, which considers both traditional Chinese medicine and modern medicine, enables a three-dimensional mapping and cross-validation of "individual innate attributes - traditional medicine - modern medicine," thereby improving the accuracy of target organ identification and ultimately enhancing the accuracy of insurance product recommendations.

[0076] This invention combines traditional Chinese medicine health risks based on birth dates and times (Bazi) with multimodal modern health information to create a comprehensive individual health profile for each insured individual. This profile identifies the corresponding risk-prone organs, enabling cross-validation and comprehensive coverage of traditional Chinese medicine and modern medicine. This improves the comprehensiveness and accuracy of the risk-prone organs, thereby enhancing the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations. This meets the growing demand for personalized insurance products and improves the effectiveness of the aforementioned technical means of insurance product recommendation.

[0077] Please see Figure 3 , Figure 3 This is a schematic diagram of the second sub-process of the AI-based insurance product recommendation processing method provided in an embodiment of the present invention. Figure 3 As shown, in this embodiment, before predicting the health risk corresponding to the individual's innate endowment attributes based on the personal innate endowment attributes and the preset Chinese Classics-Modern Medicine Information Database, and based on the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, and obtaining the estimated health risk corresponding to the preset insurance individual customer, the method further includes:

[0078] S31. Based on the preset data source of information from ancient Chinese classics and using natural language processing technology, determine the binary mapping relationship between an individual's innate endowment attributes and traditional medicine;

[0079] S32. Determine a preset medical state individual sample set, wherein the preset medical state individual sample set includes several preset medical state individual samples;

[0080] S33. Based on the same physical condition of the individual sample in the preset medical state, determine the traditional medical information data and modern medical information data corresponding to the individual sample in the preset medical state.

[0081] S34. Based on the traditional medical information data and the modern medical information data, and based on the preset medical binary relation mapping learning LSTM model, determine the medical health binary mapping relationship between the traditional medicine and modern medicine;

[0082] S35. Based on the binary mapping relationship between traditional Chinese culture and health and the binary mapping relationship between medicine and health, construct a ternary mapping relationship between "personal innate endowment attributes - traditional medicine - modern medicine" to obtain a preset traditional Chinese culture-modern medicine information database.

[0083] Interpretatively, based on a preset data source of information from classical Chinese texts, and using natural language processing technologies including but not limited to word segmentation, part-of-speech tagging, and syntactic analysis, this method automatically extracts classical Chinese health knowledge from these texts, including but not limited to personal innate attributes, constitution classification, and the association between solar terms and diseases. This knowledge is used to construct a structured knowledge base, enabling the mapping path between personal innate attributes and traditional medicine, including but not limited to "personal innate attributes → internal organs → emotions → diseases." For example, "the liver belongs to wood → anger damages the liver → risk of hypertension." This establishes a binary mapping relationship between personal innate attributes and traditional Chinese health knowledge. The preset data source of information from classical Chinese texts includes, but is not limited to, the *Huangdi Neijing*, *Shanghan Zabing Lun*, and *Nanjing*.

[0084] Furthermore, a pre-defined medical state individual sample set is established. This set includes several pre-defined medical state individual samples, including but not limited to disease incidence data, gene-chronic disease association studies, and personal health information data corresponding to historical case data. Medical state individuals refer to individuals in different health or disease states, which can be classified through dynamic assessment. Medical state individuals include, but are not limited to, healthy individuals, at-risk individuals, diseased individuals, and recovering individuals. Healthy individuals are those whose physiological, psychological, and social adaptations are all in good condition. At-risk individuals are those with abnormal biomarkers, lifestyle risks, or early functional decline but not meeting disease criteria. Diseased individuals are those diagnosed with a specific disease. Recovering individuals represent those in the stage of functional recovery after disease treatment or stable management of chronic diseases. The medical state individual samples corresponding to diseased individuals are the main information data for health risk assessment.

[0085] Furthermore, based on the same physical state of the individual samples in the preset medical state (i.e., under the same physical state corresponding to the individual samples in the preset medical state), the traditional medical information data and modern medical information data corresponding to the individual samples in the preset medical state are determined. The traditional medical information data describes the individual's physical state (especially symptoms) from the perspective of traditional medicine, while the modern medical information data describes the individual's physical state (especially symptoms) from the perspective of modern medicine. Then, based on the traditional and modern medical information data, and using a preset medical binary relation mapping to learn an LSTM model, an LSTM neural network model is used to extract corresponding features from the traditional and modern medical information data to identify the corresponding mapping relationship between traditional and modern medicine (such as blood lipids, blood sugar, blood pressure, diseases, etc.), thereby determining the medical-health binary mapping relationship between traditional and modern medicine. The preset medical binary relation mapping learning LSTM model represents a model that learns the corresponding association relationship between traditional and modern medical information data based on an LSTM model, and the medical-health binary mapping relationship represents the corresponding mapping relationship between traditional and modern medicine. Finally, based on the binary mapping relationship between traditional Chinese culture and health and the binary mapping relationship between medicine and health, a ternary mapping relationship is constructed between "individual innate endowment attributes - traditional medicine - modern medicine", resulting in a pre-set traditional Chinese culture-modern medicine information database.

[0086] This invention constructs a ternary mapping relationship between "individual innate attributes - traditional medicine - modern medicine" by determining the binary mapping relationship between traditional Chinese culture and health and the binary mapping relationship between medical health. This integrates knowledge of traditional Chinese culture and health, modern medicine, and insurance products with the user's individual innate attributes to achieve insurance product recommendations based on the correlation between "individual innate attributes - disease risk - insurance". This improves the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations, and enhances the effectiveness of the aforementioned technical means of insurance product recommendation.

[0087] In one embodiment, after pushing the target insurance product plan to the preset individual insurance customer, the method further includes:

[0088] Monitor the dynamic health information data of wearable devices corresponding to the preset insurance individual customers;

[0089] Determine whether the dynamic health information data of the wearable device meets preset health behavior conditions;

[0090] If the wearable device's dynamic health information data meets preset health behavior conditions, the premium corresponding to the target insurance product plan will be discounted according to the preset premium discount processing method.

[0091] If the dynamic health information data of the wearable device does not meet the preset health behavior conditions, the premium corresponding to the target insurance product plan will not be discounted according to the preset premium discount processing method.

[0092] Explaining this, the pre-set health behavior conditions, or preset health behavior conditions, refer to the pre-defined quantitative health indicators corresponding to the individual insured customer, including but not limited to sleep quality, exercise volume, and heart rate, meeting the pre-set criteria for determining healthy behavior. For example, preset health behavior conditions could be expressed as meeting exercise targets for 30 consecutive days, stable blood pressure, or normal heart rate. Furthermore, the pre-set premium discount processing method, or preset premium discount processing method, refers to the method of applying a corresponding discount to the premium of the insurance product corresponding to the pre-defined individual insured customer under the pre-set conditions. Pre-set premium discount processing methods include, but are not limited to, uniformly reducing the premium of different insurance products, applying different discounts to different insurance products, or gradually reducing the premium of different insurance products in a tiered manner. The preset premium discount processing method can be set according to the insurance business.

[0093] Based on the above settings, dynamic health information data of pre-defined insurance individual customers is collected from wearable devices, including but not limited to smartwatches, smart bracelets, and smartphones. This data is then monitored to determine if it meets pre-defined health behavior conditions. If the conditions are met, a discount is applied to the premium for the target insurance product plan according to a pre-defined premium discount processing method. Otherwise, if the conditions are not met, the premium remains unchanged, maintaining the original premium standard. This achieves a positive feedback loop of "behavior improvement—risk reduction—premium discount" and dynamically optimized insurance product recommendations, significantly improving the timeliness and intelligence of insurance risk management and enabling dynamic optimization of insurance product recommendations.

[0094] This invention, through wearable devices, monitors the health behaviors of pre-defined insurance clients, obtaining corresponding dynamic health information data from these devices. When the dynamic health information data meets pre-defined health behavior conditions, a discount is applied to the premiums of these clients. This provides positive incentives and guidance to these clients, creating a positive feedback loop of "improved client behavior – reduced risk – premium discount." This achieves a dynamic optimization mechanism for personal health information and insurance product recommendations based on wearable devices and health behaviors, continuously optimizing insurance product recommendations. This further improves the credibility, personalization, accuracy, timeliness, and practicality of insurance product recommendations, thereby enhancing the timeliness and intelligence of insurance risk management.

[0095] 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.

[0096] In one embodiment, an AI-based insurance product recommendation processing device is provided, which corresponds one-to-one with the AI-based insurance product recommendation processing method described in the above embodiments. Please refer to [link / reference]. Figure 4 , Figure 4 This is a schematic block diagram of an artificial intelligence-based insurance product recommendation processing device provided in an embodiment of the present invention. Figure 4 As shown, the AI-based insurance product recommendation processing device 40 includes a first determination module 41, a health risk prediction module 42, an insurance product matching module 43, a prompt word embedding module 44, an insurance plan generation module 45, and an insurance plan push module 46. Detailed descriptions of each of these functional modules are as follows:

[0097] The first determining module 41 is used to determine the innate endowment attributes corresponding to the preset insurance individual customer; the health risk prediction module 42 is used to predict the health risk corresponding to the innate endowment attributes based on the innate endowment attributes and the preset Chinese Classics-Modern Medicine Information Database, and based on the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, to obtain the estimated health risk corresponding to the preset insurance individual customer; the insurance product matching module 43 is used to match the estimated health risk with the preset insurance product knowledge graph, and based on the preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, to obtain the initial insurance product; the prompt word embedding module 44 is used to embed the initial insurance product into the preset prompt word template to obtain the target prompt word; the insurance plan generation module 45 is used to generate the target insurance product plan corresponding to the preset insurance individual customer based on the target prompt word and based on the preset insurance product plan generation module included in the preset retrieval enhancement generation scheme; the insurance plan push module 46 is used to push the target insurance product plan to the preset insurance individual customer.

[0098] In one embodiment, the first determining module 41 includes: a first determining submodule, used to determine the birth date corresponding to a preset insurance individual customer, and to determine the original innate endowment attribute corresponding to the preset insurance individual customer based on the birth date; a second determining submodule, used to determine the solar term information data and regional climate information data corresponding to the residence of the preset insurance individual customer; a third determining submodule, used to determine the solar term risk correction factor corresponding to the solar term information data, and to determine the regional climate correction factor corresponding to the regional climate information data; and a correction submodule, used to correct the original innate endowment attribute based on the solar term risk correction factor and the regional climate correction factor to obtain the innate endowment attribute corresponding to the preset insurance individual customer.

[0099] In one embodiment, the health risk prediction module 42 includes: a fourth determining submodule, used to determine the constitution type of the individual's innate endowment attribute corresponding to the preset insurance individual customer based on the individual's innate endowment attribute; a fifth determining submodule, used to determine the risk target organ corresponding to the preset insurance individual customer based on the individual's innate endowment attribute constitution type; and a health risk prediction submodule, used to predict the modern medical health risk corresponding to the risk target organ based on the preset mapping path of "individual innate endowment attribute-traditional medicine-modern medicine" and the preset health risk retrieval submodule included in the preset Chinese classics-modern medicine information database and the preset retrieval enhancement generation scheme, thereby obtaining the estimated health risk corresponding to the preset insurance individual customer.

[0100] In one embodiment, the fifth determining submodule includes: a sixth determining submodule, used to determine the TCM constitution type corresponding to the individual's innate endowment attribute constitution type; and a seventh determining submodule, used to determine the risk target organs corresponding to the preset insurance individual customer based on the TCM constitution type.

[0101] In one embodiment, the seventh determining submodule includes: an eighth determining submodule, used to determine the preset multimodal health information data corresponding to the preset insurance individual customer; a profiling submodule, used to create a health risk profile of the preset insurance individual customer based on the individual's innate constitution type, the traditional Chinese medicine constitution type, and the preset multimodal health information data, to obtain the individual customer health risk profile corresponding to the preset insurance individual customer; and a ninth determining submodule, used to determine the risk target organs corresponding to the preset insurance individual customer based on the individual customer health risk profile.

[0102] In one embodiment, the insurance product recommendation processing device 40 further includes: a second determining module, used to determine a binary mapping relationship between an individual's innate endowment attributes and traditional medicine based on a preset data source of information from ancient Chinese classics and natural language processing technology; a third determining module, used to determine a preset set of individual samples in medical states, the preset set of individual samples in medical states containing several preset individual samples in medical states; a fourth determining module, used to determine the traditional medical information data and modern medical information data corresponding to the same physical state of the individual samples in medical states; a fifth determining module, used to determine a binary mapping relationship between traditional medicine and modern medicine based on the traditional medical information data and the modern medical information data and a preset medical binary mapping learning LSTM model; and a database construction module, used to construct a ternary mapping relationship between "individual innate endowment attributes - traditional medicine - modern medicine" based on the binary mapping relationship between traditional Chinese classics and health and the binary mapping relationship between medical health, to obtain a preset traditional Chinese classics-modern medicine information database.

[0103] In one embodiment, the insurance product recommendation processing device 40 further includes: a health monitoring module for monitoring the dynamic health information data of the wearable device corresponding to the preset insurance individual customer; a judgment module for judging whether the dynamic health information data of the wearable device meets preset health behavior conditions; and a discount processing module for discounting the premium corresponding to the target insurance product plan according to a preset premium discount processing method when the dynamic health information data of the wearable device meets the preset health behavior conditions.

[0104] Specific limitations regarding the AI-based insurance product recommendation processing device can be found in the limitations of the AI-based insurance product recommendation processing method described above, and will not be repeated here. Each module in the aforementioned AI-based insurance product recommendation processing device 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, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0105] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As 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 stored 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 server-side method for recommending insurance products based on artificial intelligence.

[0106] 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 non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. 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 client-side functions or steps of an artificial intelligence-based insurance product recommendation processing method.

[0107] 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 implement the steps of the insurance product recommendation processing method described above.

[0108] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the insurance product recommendation processing method described above.

[0109] 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.

[0110] 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. When executed, the computer program 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 by this invention 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.

[0111] 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.

[0112] The software tools or components not belonging to this company that appear in the embodiments of this invention are merely illustrative examples and do not represent actual use. Furthermore, the data collection methods described in the embodiments of this invention comply with relevant laws and regulations, such as China's Personal Information Protection Law, GDPR (General Data Protection Regulation) of the European Union, or information security standards of other countries and regions.

[0113] 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 recommending insurance products based on artificial intelligence, characterized in that, include: Determine the individual innate endowment attributes corresponding to the pre-set individual insurance customers; Based on the individual's innate endowment attributes and the preset Chinese Classics-Modern Medicine Information Database, and based on the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the health risk corresponding to the individual's innate endowment attributes is predicted, and the estimated health risk corresponding to the preset insurance individual customer is obtained. Based on the knowledge graph of the estimated health risks and the preset insurance products, and based on the preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, the insurance products corresponding to the estimated health risks are matched to obtain the initial insurance products; The initial insurance product is embedded into a preset prompt template to obtain the target prompt. Based on the target prompts and the preset insurance product plan generation module included in the preset search enhancement generation scheme, a target insurance product plan corresponding to the preset individual insurance customer is generated. The target insurance product plan is pushed to the preset individual insurance customers.

2. The insurance product recommendation processing method based on artificial intelligence as described in claim 1, characterized in that, Determine the innate personal attributes of the pre-defined insurance individual customer, including: Determine the birth date corresponding to the preset insurance individual customer, and determine the original innate endowment attribute corresponding to the preset insurance individual customer based on the birth date; Determine the seasonal information data and regional climate information data corresponding to the residence of the preset insurance individual customer; Determine the solar term risk correction factor corresponding to the solar term information data, and determine the regional climate correction factor corresponding to the regional climate information data; Based on the solar term risk correction factor and the regional climate correction factor, the original individual innate endowment attributes are corrected to obtain the individual innate endowment attributes corresponding to the preset insurance individual customer.

3. The insurance product recommendation processing method based on artificial intelligence as described in claim 1, characterized in that, Based on the individual's innate endowment attributes and a preset traditional Chinese culture-modern medicine information database, and using the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the health risks corresponding to the individual's innate endowment attributes are predicted, resulting in the estimated health risks corresponding to the preset insurance individual customer, including: Based on the individual's innate endowment attributes, determine the individual's innate endowment attribute constitution type corresponding to the preset insurance individual customer; Based on the individual's innate physical constitution and attributes, the risk target organs corresponding to the preset insurance individual customer are determined; Based on the target organs and the preset mapping path of "personal innate endowment attributes - traditional medicine - modern medicine", and based on the preset Chinese classics - modern medicine information database and the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, the modern medical health risk corresponding to the target organs is predicted, and the estimated health risk corresponding to the preset insurance individual customer is obtained.

4. The insurance product recommendation processing method based on artificial intelligence as described in claim 3, characterized in that, Based on the individual's innate physical constitution and attributes, the risk target organs corresponding to the preset insurance individual customer are determined, including: Determine the TCM constitution type corresponding to the individual's innate endowment attribute constitution type; Based on the TCM constitution type, the target organs for risk are determined for the pre-set insurance individual customer.

5. The insurance product recommendation processing method based on artificial intelligence as described in claim 4, characterized in that, Based on the aforementioned TCM constitution type, the target organs for risk corresponding to the pre-set insurance individual customer are determined, including: Determine the preset multimodal health information data corresponding to the preset insurance individual customer; Based on the individual's innate constitution type, the TCM constitution type, and the preset multimodal health information data, a health risk profile is created for the preset insurance individual customer to obtain the individual customer health risk profile corresponding to the preset insurance individual customer. Based on the individual customer's health risk profile, the risk target organs corresponding to the preset insurance individual customer are determined.

6. The insurance product recommendation processing method based on artificial intelligence as described in any one of claims 1-5, characterized in that, Based on the individual's innate endowment attributes and a preset Chinese Classics-Modern Medicine information database, and using the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, before predicting the health risk corresponding to the individual's innate endowment attributes and obtaining the estimated health risk corresponding to the preset insurance individual customer, the process further includes: Based on a pre-set data source of information from ancient Chinese classics and using natural language processing technology, the binary mapping relationship between an individual's innate endowment attributes and traditional medicine is determined. Determine a preset medical state individual sample set, which includes several preset medical state individual samples; Based on the same physical condition of the individual sample in the preset medical state, determine the traditional medical information data and modern medical information data corresponding to the individual sample in the preset medical state; Based on the traditional medical information data and the modern medical information data, and using a pre-defined medical binary relation mapping to learn an LSTM model, the medical health binary mapping relationship between traditional medicine and modern medicine is determined. Based on the binary mapping relationship between traditional Chinese culture and health and the binary mapping relationship between medicine and health, a ternary mapping relationship is constructed between "individual innate endowment attributes - traditional medicine - modern medicine" to obtain a preset traditional Chinese culture-modern medicine information database.

7. The insurance product recommendation processing method based on artificial intelligence as described in claim 6, characterized in that, After pushing the target insurance product plan to the preset individual insurance customers, the process also includes: Monitor the dynamic health information data of wearable devices corresponding to the preset insurance individual customers; Determine whether the dynamic health information data of the wearable device meets preset health behavior conditions; If the dynamic health information data of the wearable device meets the preset health behavior conditions, the premium corresponding to the target insurance product plan will be discounted according to the preset premium discount processing method.

8. An insurance product recommendation processing device based on artificial intelligence, characterized in that, include: The first determining module is used to determine the innate endowment attributes of the pre-set insurance individual customers. The health risk prediction module is used to predict the health risk corresponding to the individual's innate endowment attributes based on the individual's innate endowment attributes and the preset Chinese Classics-Modern Medicine Information Database, and based on the preset health risk retrieval submodule included in the preset retrieval enhancement generation scheme, so as to obtain the estimated health risk corresponding to the preset insurance individual customer. The insurance product matching module is used to match the insurance products corresponding to the estimated health risks based on the estimated health risks and the preset insurance product knowledge graph, and based on the preset insurance product retrieval submodule included in the preset retrieval enhancement generation scheme, to obtain the initial insurance products. The prompt word embedding module is used to embed the initial insurance product into a preset prompt word template to obtain the target prompt word; The insurance plan generation module is used to generate a target insurance product plan corresponding to the preset insurance individual customer based on the target prompt words and the preset insurance product plan generation module included in the preset search enhancement generation plan. The insurance plan push module is used to push the target insurance product plan to the preset individual insurance customers.

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 artificial intelligence-based insurance product recommendation processing method 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 the processor, it implements the steps of the artificial intelligence-based insurance product recommendation processing method as described in any one of claims 1 to 7.

Images