Intelligent insurance sales all-in-one machine and system thereof

Abstract

The invention discloses an intelligent insurance sales all-in-one machine and a system thereof, and relates to the technical field of insurance sales, the intelligent insurance sales all-in-one machine comprises an insurance sales all-in-one machine body, the bottom surface of the insurance sales all-in-one machine body is provided with a group of self-locking universal wheels, and the front surface of the insurance sales all-in-one machine body is provided with a display screen. The heat dissipation effect of the insurance sales all-in-one machine body can be guaranteed, the service life of the insurance sales all-in-one machine body can be effectively prolonged, personal information, risk preference and financial conditions of clients can be accurately mastered through the user information collection module, a solid foundation is provided for personalized recommendation, and the user experience is improved. The personalized insurance product recommendation module applies an innovative multi-dimensional feature fusion algorithm, can comprehensively evaluate personal information, risk preference, financial conditions and the like of the user, provides personalized insurance product recommendation for the user, continuously optimizes a recommendation model, improves the accuracy and pertinence of recommendation, and meets different requirements of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of insurance sales, and in particular to an intelligent all-in-one insurance sales machine and a system thereof. Background Art

[0002] Insurance, as an important financial tool, aims to provide people with risk protection. It organizes many individuals facing the same risks to share possible losses. There are many types of insurance, covering life insurance, health insurance, property insurance and other fields, to meet the risk protection needs of different groups of people in different scenarios. The insurance sales machine is an intelligent device that integrates advanced technology. It is usually equipped with a high-resolution display, input device and powerful data processing capabilities. Its main function is to integrate various aspects of insurance sales and provide customers with a one-stop insurance service experience. On the one hand, it can efficiently collect customers' personal information, including age, gender, occupation, income, family status, etc., to provide a data basis for subsequent insurance product recommendations. On the other hand, it displays detailed information of various insurance products through an intuitive interface, such as coverage, premiums, claims conditions, etc., to help customers better understand and compare different products. In addition, the insurance sales machine can also realize fast transaction processing, such as filling out applications online, printing contracts instantly, etc., to improve sales efficiency and customer satisfaction.

[0003] The insurance sales machine is a device that integrates hardware and software systems. Its function is to provide customers with convenient insurance sales services. It can effectively solve the problem of low efficiency of current insurance sales relying on manual methods. However, the existing insurance sales machines still have certain shortcomings when used. In terms of hardware, the heat dissipation capacity of the existing insurance sales machines is insufficient. The heat of the insurance sales machines cannot be dissipated in time, which will affect the service life of the insurance sales machines. In addition, the existing insurance sales machines are not intelligent enough in terms of systems, and it is difficult to fully analyze user information, resulting in the recommended insurance products often not accurately matching customer needs, and failing to make full use of big data and artificial intelligence technology to provide customers with personalized insurance product recommendations and high-quality services. Therefore, it is necessary to propose an intelligent insurance sales machine and its system to solve the problems in the existing technology. Summary of the invention

[0004] The purpose of the present invention is to make up for the shortcomings of the prior art and to provide an intelligent insurance sales machine and its system, which can ensure the heat dissipation effect of the insurance sales machine body, effectively extend the service life of the insurance sales machine, and comprehensively evaluate the user's personal information, risk preference, financial status, etc., to provide users with personalized insurance product recommendations, continuously optimize the recommendation model, improve the accuracy and pertinence of the recommendations, and meet the different needs of users.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent insurance sales machine, comprising an insurance sales machine body, a group of self-locking universal wheels are installed on the bottom surface of the insurance sales machine body, a display screen is installed on the front of the insurance sales machine body, a touch screen and an identity recognition panel are installed on the upper surface of the insurance sales machine body, an insurance policy printing outlet is arranged on the front of the insurance sales machine body, an insurance policy storage slot is installed on the front of the insurance sales machine body, an air intake assembly is installed on the left side of the insurance sales machine body, the air intake assembly includes an air intake frame fixedly embedded in the left side of the insurance sales machine body, a dustproof net is installed on the inner wall of the air intake frame, and a heat dissipation assembly is installed on the right side of the insurance sales machine body.

[0006] Furthermore, the heat dissipation assembly includes an air outlet duct, and a group of mounting ears are installed on the outer surface of the air outlet duct, and each of the mounting ears is connected to the body of the insurance sales machine through a mounting bolt.

[0007] Furthermore, a protective net is installed on the inner wall of the air outlet duct, an exhaust fan is arranged inside the air outlet duct, and the exhaust fan is connected to the inner wall of the air outlet duct through two connecting blocks.

[0008] An intelligent insurance sales machine system, the system includes: a user information collection module, a personalized insurance product recommendation module, a recommendation model optimization module and an intelligent consulting service module; The user information collection module is used to comprehensively collect the user's personal information, risk preference and financial status data to provide accurate basic data for subsequent personalized insurance product recommendations. The personalized insurance product recommendation module uses an innovative multi-dimensional feature fusion algorithm to provide users with highly personalized insurance product recommendations based on the user features acquired by the user information collection module and the insurance product features extracted by the insurance product feature extraction unit. The recommendation model optimization module is used to continuously improve the accuracy and adaptability of recommendations. The recommendation model optimization module includes: a user feedback analysis unit, a model adjustment unit and a new data fusion unit. The intelligent consulting service module is used to provide users with intelligent voice consulting functions using natural language processing technology; The personalized insurance product recommendation module includes: a user feature extraction unit, an insurance product feature extraction unit and a matching calculation unit. The user feature extraction unit extracts key features from the collected user personal information, risk preference and financial status data to provide a basis for subsequent recommendation calculations. The insurance product feature extraction unit extracts the features of various insurance products, including coverage, premiums, and claim conditions. The matching calculation unit uses a specific algorithm to perform matching calculations based on user features and insurance product features to find the insurance product that best meets user needs. The personalized insurance product recommendation module uses a unique multi-dimensional fusion recommendation algorithm, and the algorithm formula is: ,in Indicates the comprehensive score of recommended insurance products. is the number of user basic feature dimensions, is the weight coefficient of each user's basic feature dimension, User characteristics and insurance product characteristics In the The fusion function on the dimension of user basic features, is the number of user interaction feature dimensions, is the weight coefficient of each user interaction feature dimension, User Features , Insurance Product Features and user-system interaction characteristics In the The fusion function on the dimension of user interaction features, To adjust the interaction feature coefficient, the user information collection module comprehensively collects user personal information, risk preferences and financial status to form a user feature vector The insurance product feature extraction unit extracts the coverage, premium, and claim conditions of the insurance product to form an insurance product feature vector , obtain the user interaction feature vector through the interaction record between the user and the system , by continuously adjusting the weight coefficients and optimizing the fusion function, highly personalized insurance product recommendations can be achieved.

[0009] Furthermore, the user feature extraction unit in the personalized insurance product recommendation module performs innovative segmentation assignment on the user age feature, defining the user age as ,when When , the characteristic value is assigned , and further adjust the characteristic values ​​according to different small ranges within the age group. When , the characteristic value is assigned , the same refinement is performed, when When , the characteristic value is assigned And refine, for occupational characteristics, classification and assignment are carried out according to the innovative risk assessment index system of occupations, and high-risk occupations are assigned higher characteristic values ​​based on the comprehensive calculation of multiple risk sub-indicators. , low-risk occupations are assigned lower characteristic values .

[0010] Furthermore, the insurance product feature extraction unit in the personalized insurance product recommendation module adopts a new quantitative method for the insurance product coverage, defining the coverage to include multiple dimensions, including health, accident and property, and each dimension is evaluated based on an innovative protection importance evaluation function. Determine the weight value, where To protect the relevant parameters of the dimension, the premium characteristics are dynamically adjusted according to the type of insurance product and market conditions. Multiplying the basic premium value, the claim conditions are quantified according to the strictness and innovative customer adaptability evaluation indicators .

[0011] Furthermore, the fusion function in the personalized insurance product recommendation module and A feature fusion network structure based on deep learning is adopted. The network consists of multiple levels of neurons. The network parameters are continuously optimized through training so that the fusion function can accurately reflect the complex relationship between user characteristics, insurance product characteristics and user interaction characteristics.

[0012] Furthermore, the user feedback analysis unit in the recommendation model optimization module, after collecting user feedback information, adopts an innovative feedback classification algorithm to classify the reasons for non-purchase into multiple categories. For each category, based on the specific content and frequency of the feedback, the impact factor of the category is determined through complex text analysis and statistical methods. , so as to more accurately understand users’ views on recommendation results and where their needs are not met.

[0013] Furthermore, the model adjustment unit in the recommendation model optimization module adjusts the recommendation model according to the result of the user feedback analysis unit, and defines the weight of a factor in the recommendation model as , the adjusted weight is , the adjustment formula is ,in is the number of feedback categories, is the adjustment factor determined according to the importance of each category of feedback, is the impact factor of the corresponding category.

[0014] Furthermore, when a new data source is introduced into the new data fusion unit in the recommendation model optimization module, a fusion strategy based on dynamic weight allocation is adopted to define the new data as , the original recommended model is The fused model is , the fusion formula is ,in is the number of new data sources, The dynamic weight coefficient of each new data source is determined based on multiple indicators of the reliability and timeliness of the new data source. is the retention coefficient of the original recommended model.

[0015] Compared with the existing technology, the intelligent insurance sales integrated machine and its system have the following beneficial effects: 1. The present invention, through the coordinated arrangement of the insurance sales machine body, the air intake assembly and the heat dissipation assembly, can extract the air inside the insurance sales machine body by utilizing the heat dissipation assembly, so that the external air can enter the inside of the insurance sales machine body through the air intake assembly, and can conveniently replace the air inside the insurance sales machine body, thereby conveniently dissipating the heat of the insurance sales machine body, and can ensure the heat dissipation effect of the insurance sales machine body, thereby effectively extending the service life of the insurance sales machine body.

[0016] 2. The present invention can accurately grasp the customer's personal information, risk preference and financial status through the user information collection module, providing a solid foundation for personalized recommendations. The personalized insurance product recommendation module uses an innovative multi-dimensional feature fusion algorithm to provide customers with insurance product recommendations that are highly in line with their needs. The recommendation model optimization module continuously adjusts and integrates according to user feedback and new data to ensure that the recommendation results always maintain accuracy and adaptability, so that the insurance sales machine can conduct a comprehensive assessment of the user's personal information, risk preference, and financial status, provide users with personalized insurance product recommendations, continuously optimize the recommendation model, improve the accuracy and pertinence of recommendations, and meet the different needs of users.

[0017] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 It is a three-dimensional structural diagram of the right view of the intelligent insurance sales machine; Figure 2 It is a three-dimensional structural diagram of the left view of the intelligent insurance sales machine; Figure 3 This is a schematic diagram of the internal structure of the air outlet duct in the smart insurance sales machine; Figure 4 This is a schematic diagram of the structure of the system of the intelligent insurance sales machine.

[0020] In the figure: 1. Insurance sales machine body; 2. Self-locking universal wheels; 3. Display screen; 4. Touch screen; 5. Identity recognition panel; 6. Insurance policy printing outlet; 7. Insurance policy storage slot; 8. Air inlet assembly; 801. Air inlet frame; 802. Dustproof net; 9. Heat dissipation assembly; 901. Air outlet duct; 902. Mounting ears; 903. Mounting bolts; 904. Protective net; 905. Exhaust fan; 906. Connecting block. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0022] A large shopping mall has set up an intelligent insurance sales machine to provide insurance sales services to mall customers and surrounding residents. The mall is located in a bustling commercial center with a large flow of people and a diverse customer base, including office workers, housewives, and the elderly.

[0023] A 35-year-old office worker, Mr. Li, came to the all-in-one machine. The machine first guided Mr. Li to register and fill in information with a simple and clear interface. Mr. Li easily entered his personal information through the touch screen, including age, occupation (software engineer), income, family status (married, with one child), etc. At the same time, the system cooperated with a third-party credit data platform to securely obtain Mr. Li's credit record, consumption behavior and other data.

[0024] In the personal information collection module, the personal information collection unit processes Mr. Li's age information. Since Mr. Li is 35 years old and in the middle-aged stage, according to the assignment rule in the claim, an age characteristic value is assigned. , and further refine and adjust the characteristic values ​​according to the specific circumstances of the middle-aged age group. For example, considering that Mr. Li is in the rising period of his career, he has high work pressure and may face some health risks, the age characteristic value is fine-tuned. For occupational characteristics, the risk preference assessment unit analyzes the characteristics of the software engineer profession, based on the occupational innovation risk assessment index system, comprehensively considers factors such as work pressure, work environment, and career development prospects, and assigns Mr. Li an occupational characteristic value (Assuming that the occupation is assessed as a medium-risk occupation), the financial status analysis unit analyzes Mr. Li’s income, liabilities, savings and other financial data to provide an accurate basis for subsequent recommendations. At the same time, considering that Mr. Li has family burdens, his family status will also affect the eigenvalue.

[0025] The personalized insurance product recommendation module starts working, and the user feature extraction unit forms a user feature vector based on Mr. Li’s information. ,For example It may include age characteristic values, occupation characteristic values, income level, family status characteristic values, etc.

[0026] The insurance product feature extraction unit extracts the features of various insurance products provided by the cooperating insurance companies in the mall to form an insurance product feature vector For example, a major disease insurance product covers a variety of common major diseases. According to the new quantitative method, each protection dimension is evaluated according to the innovative protection importance function. Determine the weight value, assuming that the weight value of the health protection dimension is Taking into account Mr. Li’s occupation and age, the weight of the health insurance dimension is appropriately adjusted, and the premium is 500 yuan per month. The dynamic adjustment coefficient based on the premium characteristics After processing, the premium characteristic value is obtained, and the claim conditions are quantified and assigned according to the strictness and innovative customer adaptability evaluation indicators.

[0027] The recommendation algorithm starts to calculate, using a multi-dimensional fusion recommendation algorithm, the formula is ,in, and are the number of user basic feature dimensions and the number of user interaction feature dimensions respectively. Assuming that in the current calculation, , corresponding to the four basic characteristic dimensions of age, occupation, income, and family status; , corresponding to the three interactive feature dimensions of recent query times, dwell time, and clicked product types, and the weight coefficient of each dimension and The initial value is set based on experience and is adjusted as the system continues to learn and optimize. and The deep learning-based feature fusion network structure is adopted. Through training on a large amount of historical data, the network parameters are continuously optimized so that the fusion function can accurately reflect the complex relationship between user characteristics, insurance product characteristics and user interaction characteristics. During Mr. Li's interaction with the all-in-one machine, the system records his query times, stay time, clicked product types and other interaction feature vectors. Through calculation, we recommended several suitable insurance products to Mr. Li, such as major disease insurance, accident insurance, children's education fund insurance, etc.

[0028] Mr. Li viewed the recommended insurance product but did not purchase it immediately. The recommendation model optimization module began to work. The user feedback analysis unit collected Mr. Li's feedback information and classified the reasons why Mr. Li did not purchase it through an innovative feedback classification algorithm. Assuming that Mr. Li thought the premium was high, the system classified this feedback into the overpriced category and determined the influencing factors of this category based on complex text analysis and statistical methods. ,At the same time, we analyze Mr. Li’s behavioral data in the process of viewing ,products, such as the characteristics of products that he stayed for a long time but did not ,purchase, to further refine the feedback classification.

[0029] The model adjustment unit adjusts the recommendation model according to the feedback analysis results. Suppose the premium weight of a certain insurance product in the recommendation model is , the adjusted weight is , the adjustment formula is ,in, is the number of feedback categories. Assuming that there are currently 4 types of feedback, including too high price, inappropriate coverage, lack of understanding of the product, and other reasons, the adjustment coefficient is determined based on the importance of the too high price feedback category. , combined with the impact factor Adjust the premium weights and, in response to feedback on inappropriate coverage, analyze specific protection needs and adjust the weights of corresponding product features.

[0030] The new data fusion unit continuously introduces new data sources, such as the latest insurance product information launched on the market, industry claims trends, the impact of economic changes on insurance demand, etc., and adopts a fusion strategy based on dynamic weight allocation. , the original recommended model is , after fusion ,in, is the number of new data sources. Assume that there are 3 new data sources. It is the dynamic weight coefficient of each new data source, which is determined based on multiple indicators such as the reliability, timeliness, and relevance of the new data source to the current user group. It is the retention coefficient of the original recommendation model. By continuously integrating new data, the recommendation model becomes more accurate and adaptable to market changes.

[0031] Mr. Li had some questions about a recommended accident insurance and asked a question by voice: "What are the claim conditions for this accident insurance?" The intelligent consulting service module uses natural language processing technology to first perform voice recognition and convert it into text, and then extract the key information of the question through lexical analysis, syntactic analysis and other processing. Based on the key information, it searches the knowledge base to find the claim conditions for the accident insurance and answers Mr. Li in clear language. At the same time, based on Mr. Li's question, it further explores the user's needs and provides Mr. Li with some suggestions on accident insurance selection. For example, considering Mr. Li's professional characteristics and family situation, it recommends an accident insurance that includes occupational accident protection and family responsibility protection.

[0032] If Mr. Li has any questions about the terms and conditions of the insurance product, premium calculation, etc., the intelligent consulting service module can also accurately answer them and help Mr. Li better understand them by giving examples. At the same time, the system can further optimize the user feature vector based on Mr. Li’s question records to improve the accuracy of recommendations.

[0033] The effects brought by this embodiment are as follows: For user Mr. Li, through the smart insurance sales machine and its system, he received personalized insurance product recommendations that better suited his actual needs and risk tolerance. During the consultation process, he received professional and accurate answers, which enhanced his understanding and trust in insurance products. At the same time, the interactivity and convenience of the system allowed him to easily learn about insurance products while shopping in the mall, saving time.

[0034] For shopping malls and insurance companies, the installation of smart all-in-one machines has expanded the channels and coverage of insurance sales and improved sales efficiency. Through continuously optimized recommendation models, they can better meet customer needs, improve customer satisfaction and purchase conversion rates, thereby achieving increased commercial value. Shopping malls can also provide more value-added services and attract more customers through cooperation with insurance companies.

[0035] From the perspective of the overall insurance sales industry, this innovative intelligent insurance sales model has brought new development ideas and technical means to the industry, promoted the intelligent and personalized development of insurance sales, and helped to improve the service level and competitiveness of the entire industry. At the same time, the application of big data analysis and artificial intelligence algorithms has also provided more data support for the design and pricing of insurance products, promoting innovation and development in the insurance industry. Example

[0036] In a busy bank branch, an intelligent insurance sales machine has been set up to provide convenient insurance sales services to bank customers and expand business channels for the bank.

[0037] A 45-year-old business owner, Ms. Wang, came to the bank to handle business and was guided by bank staff to the smart insurance sales machine. The machine first prompted Ms. Wang to verify her identity, and confirmed her identity through facial recognition and bank card information reading. At the same time, the system connected to the bank database to obtain information such as Ms. Wang's asset status, loan status, and financial management preferences.

[0038] In the personal information collection module, the personal information collection unit processes Ms. Tu's age information. Since Ms. Tu is 45 years old and in the middle-aged stage, according to the assignment rule in the claim, the age characteristic value is assigned. (Assuming that the characteristic values ​​of different age groups are further subdivided in the middle-aged stage), for occupational characteristics, the risk preference assessment unit assigns Ms. Wang an occupational characteristic value based on the occupational characteristics of the business owner, taking into account factors such as business risks and income fluctuations. (Medium to high risk occupation), the financial status analysis unit analyzes Ms. Wang’s assets and liabilities, income flow and other financial data in the bank to provide an accurate basis for subsequent recommendations.

[0039] The personalized insurance product recommendation module starts working, and the user feature extraction unit forms a user feature vector based on Ms. Wang’s information. , including age feature values, occupation feature values, asset status feature values, etc. The insurance product feature extraction unit extracts the features of various insurance products provided by the cooperative insurance company to form an insurance product feature vector For example, a high-end medical insurance product covers high-quality medical resources at home and abroad. According to the new quantitative method, each protection dimension is evaluated according to the innovative protection importance function. Determine the weight value, assuming that the weight value of the high-end medical resource guarantee dimension is The premium is 20,000 yuan per year, and the dynamic adjustment coefficient is based on the premium characteristics After processing, the premium characteristic value is obtained, and the claim conditions are quantified and assigned according to the strictness and innovative customer adaptability evaluation indicators. .

[0040] The recommendation algorithm starts to calculate, using a multi-dimensional fusion recommendation algorithm, the formula is ,in, and are the number of user basic feature dimensions and the number of user interaction feature dimensions respectively. Assuming that in the current calculation, , corresponding to the five basic characteristic dimensions of age, occupation, asset status, financial management preference, and family status (married, with independent children); , corresponding to the four interactive feature dimensions of recent query times, dwell time, clicked product types, and number of exchanges with bank staff, the weight coefficient of each dimension and The initial value is set based on experience and is adjusted as the system continues to learn and optimize. and The deep learning-based feature fusion network structure is adopted. Through training on a large amount of historical data, the network parameters are continuously optimized so that the fusion function can accurately reflect the complex relationship between user characteristics, insurance product characteristics and user interaction characteristics. During Ms. Wang's interaction with the all-in-one machine, the system records her interaction feature vectors such as the number of queries, dwell time, types of products clicked, and the number of exchanges with bank staff. Through calculation, we recommended several suitable insurance products to Ms. Wang, such as high-end medical insurance, corporate property insurance, and whole life insurance.

[0041] Ms. Wang checked the recommended insurance products, but expressed some concerns about the premiums. The recommendation model optimization module began to play a role. The user feedback analysis unit collected Ms. Wang’s feedback information and classified the reason for Ms. Wang’s non-purchase into the category of too high price through an innovative feedback classification algorithm. The influencing factors of this category were determined based on complex text analysis and statistical methods. At the same time, we analyzed Ms. Wang’s behavioral data during the product review process, such as her attention to the scope of protection, to further refine the feedback classification.

[0042] The model adjustment unit adjusts the recommendation model according to the feedback analysis results. Suppose the premium weight of a certain insurance product in the recommendation model is , the adjusted weight is , the adjustment formula is ,in, is the number of feedback categories. Assuming that there are currently 4 types of feedback, including too high price, inappropriate coverage, lack of understanding of the product, and other reasons, the adjustment coefficient is determined based on the importance of the too high price feedback category. , combined with the impact factor Adjust the premium weights and, in response to feedback on inappropriate coverage, analyze specific protection needs and adjust the weights of corresponding product features.

[0043] The new data fusion unit continuously introduces new data sources, such as the latest insurance product information launched on the market, industry claims trends, the impact of the macroeconomic situation on insurance demand, etc., and adopts a fusion strategy based on dynamic weight allocation. , the original recommended model is The fused model is ,in, is the number of new data sources. Assume that there are 3 new data sources. It is the dynamic weight coefficient of each new data source, which is determined based on multiple indicators such as the reliability, timeliness, and relevance of the new data source to the current user group. It is the retention coefficient of the original recommendation model. By continuously integrating new data, the recommendation model becomes more accurate and adaptable to market changes.

[0044] Ms. Wang had some questions about the recommended whole life insurance and asked the question by voice, “How is the return on this whole life insurance guaranteed?” The intelligent consulting service module uses natural language processing technology to first perform voice recognition and convert it into text. It then extracts the key information of the question through lexical analysis, syntactic analysis and other processing. It then searches the knowledge base based on the key information to find the return guarantee mechanism of the whole life insurance and answers Ms. Wang in clear language. At the same time, it further explores the user's needs based on Ms. Wang's question and provides Ms. Wang with some suggestions on asset inheritance and wealth planning.

[0045] If Ms. Wang has any questions about the terms and conditions of the insurance product, the claims process, etc., the intelligent consulting service module can also accurately answer them and help Ms. Wang better understand through case analysis and other methods. At the same time, the system can further optimize the user feature vector based on Ms. Wang’s question records to improve the accuracy of recommendations.

[0046] The effects brought by this embodiment are as follows: For user Ms. Wang, through the smart insurance sales machine and its system, she received personalized insurance product recommendations while handling business at the bank. These recommendations are more in line with her actual needs and risk tolerance. During the consultation process, she received professional and accurate answers, which enhanced her understanding and trust in insurance products. At the same time, the convenience of the system enables her to understand insurance products efficiently, providing more options for her wealth management and risk protection.

[0047] For banks, the setting of intelligent all-in-one machines has expanded business channels, provided customers with more value-added services, and enhanced customer stickiness. Through cooperation with insurance companies, banks can obtain certain intermediary business income and improve overall profitability. At the same time, the system's data analysis and recommendation functions also help banks better understand customer needs and provide customers with more personalized financial services.

[0048] From the perspective of the overall insurance sales industry, this model of setting up integrated smart insurance sales machines at bank outlets provides new channels and methods for insurance sales. The application of big data analysis and artificial intelligence algorithms has improved the efficiency and accuracy of insurance sales, promoted the intelligent and personalized development of insurance sales, and helped to improve the service level and competitiveness of the entire industry.

[0049] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the same elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. Intelligent insurance sales machine, characterized by: The invention comprises an integrated insurance sales machine body (1), a group of self-locking universal wheels (2) are installed on the bottom surface of the integrated insurance sales machine body (1), a display screen (3) is installed on the front surface of the integrated insurance sales machine body (1), a touch screen (4) and an identity recognition panel (5) are installed on the upper surface of the integrated insurance sales machine body (1), an insurance policy printing outlet (6) is arranged on the front surface of the integrated insurance sales machine body (1), an insurance policy storage slot (7) is installed on the front surface of the integrated insurance sales machine body (1), an air intake assembly (8) is installed on the left side of the integrated insurance sales machine body (1), the air intake assembly (8) comprises an air intake frame (801) fixedly embedded on the left side of the integrated insurance sales machine body (1), a dustproof net (802) is installed on the inner wall of the air intake frame (801), and a heat dissipation assembly (9) is installed on the right side of the integrated insurance sales machine body (1).

2. The smart insurance sales machine according to claim 1, characterized in that: The heat dissipation assembly (9) comprises an air outlet pipe (901), and a group of mounting ears (902) are installed on the outer surface of the air outlet pipe (901), and each of the mounting ears (902) is connected to the insurance sales machine body (1) via a mounting bolt (903).

3. The intelligent insurance sales machine according to claim 2, characterized in that: A protective net (904) is installed on the inner wall of the air outlet pipe (901), an exhaust fan (905) is arranged inside the air outlet pipe (901), and the exhaust fan (905) is connected to the inner wall of the air outlet pipe (901) via two connecting blocks (906).

4. The system of the intelligent insurance sales machine is characterized by: The system includes: user information collection module, personalized insurance product recommendation module, recommendation model optimization module and intelligent consulting service module; The user information collection module is used to comprehensively collect the user's personal information, risk preference and financial status data to provide accurate basic data for subsequent personalized insurance product recommendations. The personalized insurance product recommendation module uses an innovative multi-dimensional feature fusion algorithm to provide users with highly personalized insurance product recommendations based on the user features acquired by the user information collection module and the insurance product features extracted by the insurance product feature extraction unit. The recommendation model optimization module is used to continuously improve the accuracy and adaptability of recommendations. The recommendation model optimization module includes: a user feedback analysis unit, a model adjustment unit and a new data fusion unit. The intelligent consulting service module is used to provide users with intelligent voice consulting functions using natural language processing technology; The personalized insurance product recommendation module includes: a user feature extraction unit, an insurance product feature extraction unit and a matching calculation unit. The user feature extraction unit extracts key features from the collected user personal information, risk preference and financial status data to provide a basis for subsequent recommendation calculations. The insurance product feature extraction unit extracts the features of various insurance products, including coverage, premiums, and claim conditions. The matching calculation unit uses a specific algorithm to perform matching calculations based on user features and insurance product features to find the insurance product that best meets user needs. The personalized insurance product recommendation module uses a unique multi-dimensional fusion recommendation algorithm, and the algorithm formula is: ,in Indicates the comprehensive score of recommended insurance products. is the number of user basic feature dimensions, is the weight coefficient of each user's basic feature dimension, User Features and insurance product features In the The fusion function on the dimension of user basic features, is the number of user interaction feature dimensions, is the weight coefficient of each user interaction feature dimension, User Features , Insurance Product Features and user-system interaction characteristics In the The fusion function on the dimension of user interaction features, To adjust the interaction feature coefficient, the user information collection module comprehensively collects user personal information, risk preferences and financial status to form a user feature vector The insurance product feature extraction unit extracts the coverage, premium, and claim conditions of the insurance product to form an insurance product feature vector , obtain the user interaction feature vector through the interaction record between the user and the system , by continuously adjusting the weight coefficients and optimizing the fusion function, highly personalized insurance product recommendations can be achieved.

5. The system of the intelligent insurance sales machine according to claim 4 is characterized in that: The user feature extraction unit in the personalized insurance product recommendation module performs innovative segmentation assignment on the user age feature and defines the user age as ,when When , the characteristic value is assigned , and further adjust the characteristic values ​​according to different small ranges within the age group. When , the characteristic value is assigned , the same refinement is performed, when When , the characteristic value is assigned And refine, for occupational characteristics, classification and assignment are carried out according to the innovative risk assessment index system of occupations, and high-risk occupations are assigned higher characteristic values ​​based on the comprehensive calculation of multiple risk sub-indicators. , low-risk occupations are assigned lower characteristic values .

6. The system of the intelligent insurance sales machine according to claim 4, characterized in that: The insurance product feature extraction unit in the personalized insurance product recommendation module adopts a new quantitative method for the insurance product coverage, defining the coverage to include multiple dimensions, including health, accident and property, and each dimension is evaluated based on an innovative protection importance evaluation function. Determine the weight value, where To protect the relevant parameters of the dimension, the premium characteristics are dynamically adjusted according to the type of insurance product and market conditions. Multiplying the basic premium value, the claim conditions are quantified according to the strictness and innovative customer adaptability evaluation indicators .

7. The system of the intelligent insurance sales machine according to claim 4, characterized in that: The fusion function in the personalized insurance product recommendation module and A feature fusion network structure based on deep learning is adopted. The network consists of multiple levels of neurons. The network parameters are continuously optimized through training so that the fusion function can accurately reflect the complex relationship between user characteristics, insurance product characteristics and user interaction characteristics.

8. The system of the intelligent insurance sales machine according to claim 4, characterized in that: The user feedback analysis unit in the recommendation model optimization module collects user feedback information and uses an innovative feedback classification algorithm to classify the reasons for non-purchase into multiple categories. For each category, the impact factor of the category is determined through complex text analysis and statistical methods based on the specific content and frequency of the feedback. , so as to more accurately understand users’ views on recommendation results and where their needs are not met.

9. The system of the smart insurance sales machine according to claim 4, characterized in that: The model adjustment unit in the recommendation model optimization module adjusts the recommendation model according to the result of the user feedback analysis unit, and defines the weight of a factor in the recommendation model as , the adjusted weight is , the adjustment formula is ,in is the number of feedback categories, is the adjustment factor determined according to the importance of each category of feedback. is the impact factor of the corresponding category.

10. The system of the intelligent insurance sales machine according to claim 4, characterized in that: When introducing a new data source, the new data fusion unit in the recommendation model optimization module adopts a fusion strategy based on dynamic weight allocation and defines the new data as , the original recommended model is The fused model is , the fusion formula is ,in is the number of new data sources, The dynamic weight coefficient of each new data source is determined based on multiple indicators of the reliability and timeliness of the new data source. is the retention coefficient of the original recommended model.

Images