TWM625393U - Personalized recommendation device for insurance products - Google Patents

Abstract

The present invention discloses a personalized recommendation device for insurance products, which includes a training device and a running device. The training device provides a plurality of modules to be trained, and is used for training the above-mentioned plurality of modules to be trained according to the user's historical behavior information and insurance product information, so as to obtain a plurality of training modules and a plurality of trained data, and Stored in the data and model database. The running module signal is connected to the training module for selecting at least one training model and at least one training data from the plurality of trained modules and the plurality of trained data stored in the data and model database according to the requirements of the recommended algorithm, and The selected at least one trained module is run based on the selected at least one trained data and the input information, thereby generating at least one multi-level insurance recommendation result.

Description

Personalized recommendation device for insurance products

The invention relates to a device for personalizing insurance products, in particular to a device for recommending insurance products, insurance product types, payment item amounts and insurance products suitable for consumers with multi-level insurance recommendation results based on user behavior. device.

Nowadays, the social structure is changing, and life risks are everywhere, which makes the importance of insurance even more important. However, most of the current insurance product recommendations are made by consumers who actively contact the business specialists of various insurance companies, or use advance information based on ethnic groups, etc. The classified business recommendation combination or the policy combination selected by most people, but consumers cannot fully understand the recommended content of the insurance product, and then make the choice of the insurance product.

Another way to recommend insurance products is to guide consumers to input various information to correspond to their needs, such as gender, age, occupation, and design contextual questionnaires, etc., thereby recommending at least one insurance product that corresponds to consumers’ needs. Suggest the amount of its insurance benefit items for consumers to choose. Although this method of recommending insurance products can guide consumers to understand their own needs as much as possible, it still cannot make consumers understand the relationship between the order of recommended combinations and consumers’ needs, and consumers still cannot choose the most suitable insurance products for their needs. .

The purpose of the present invention is to provide a personalized recommendation device for insurance products, which outputs at least one multi-level insurance recommendation result according to the received recommendation algorithm requirements and input information, wherein the personalized recommendation device for insurance products includes a training device and Run the mod. The training device provides a plurality of modules to be trained, and is used to train the plurality of modules to be trained according to the user's historical behavior information and insurance product information, so as to obtain a plurality of trained models and a plurality of trained data, and store them in the Data and Model Database. The running module signal is connected to the training device for selecting at least one training model and selecting at least one training data from a plurality of trained modules and a plurality of trained data stored in the data and model database according to the requirements of the recommended algorithm, and run the selected at least one trained module based on the selected at least one trained data and the input information, thereby generating the above-mentioned multi-level insurance recommendation result. The above-mentioned multi-level insurance recommendation results are the first-level recommendation results and the X-th level recommendation results, where X is greater than or equal to 2.

Optionally, the above-mentioned multi-level insurance recommendation result further includes the J-th level recommendation result, where J is not equal to X.

Optionally, the above-mentioned first-level recommendation result includes at least one recommended insurance type, the second-level recommendation result includes at least one recommended insurance product type of the above-mentioned recommended insurance type, and the third-level recommendation result is the recommended insurance in the above-mentioned recommended insurance types. The amount of at least one claim payment item of the commodity type.

Optionally, the above-mentioned multi-level recommendation result further includes a fourth-level recommendation result, and the fourth-level recommendation result is at least one recommended insurance product.

Optionally, the above-mentioned personalized recommendation device for insurance products further includes a user behavior collection device for collecting user behavior through the user device, and the user behavior history information is determined by the user behavior.

Optionally, a plurality of users have a unique identification and are associated with the user's historical behavior information.

Optionally, the above-mentioned personalized recommendation device for insurance products further includes a front-end device signal connected to the running device for receiving recommendation algorithm requirements and input information.

Optionally, the above-mentioned personalized recommendation device for insurance products further includes the data and model database, and the signal connects the running device and the training device.

Optionally, the above-mentioned personalized recommendation device for insurance products further includes a user history behavior database signal connected to the training device for storing user history information, and transmitting the user behavior history information to the training device and insurance product data. The library, the signal is connected to the training device, is used to store the information of a plurality of insurance products.

Based on the above, the personalization recommendation device for insurance products of the present invention allows users to select insurance products according to their personal needs, and can better understand the relationship between the order of recommended combinations and their own needs for users to choose. The best insurance product for your needs.

In order to make those skilled in the art further understand the technical features, contents and advantages of the novel creation and the effects that can be achieved, the following describes the contents of the novel in detail with appropriate embodiments in conjunction with the representations of the drawings. It is only for the purpose of illustrating and assisting the description of the present invention, and is not intended to limit the scope of rights of the present invention in the actual embodiments.

In view of the problems of the prior art, the present invention mainly provides a personalized recommendation device for insurance products according to the user's own needs, and enables the user to understand the relationship between the order of the recommended combination and his own needs, so that the user can Choose the insurance product that best suits your needs.

Please refer to FIG. 1 . FIG. 1 is a functional block diagram of a personalization recommendation device 300 for insurance products according to an embodiment of the present invention. The above-mentioned personalized recommendation device 300 for insurance products outputs at least one multi-level insurance recommendation result according to the received recommendation algorithm requirements and input information. The above-mentioned personalized recommendation device 300 for insurance products includes a training device 310 and an operation device 310. . The above-mentioned training device 320 is provided with a plurality of modules to be trained, and is used for training the above-mentioned plurality of modules to be trained according to the user's historical behavior information and insurance product information, so as to obtain a plurality of trained models and a plurality of trained models. The training data is stored in the data and model database 330 . The operating device 310 is connected to the training device 320 by a signal for selecting at least one of the above-mentioned plurality of trained modules and the above-mentioned plurality of trained data stored in the above-mentioned data and model database 330 according to the above-mentioned recommendation algorithm requirements The model is trained and at least one trained data is selected, and the selected at least one trained module is run based on the selected at least one trained data and the above-mentioned input information, thereby generating the above-mentioned multi-level insurance recommendation result.

In addition, the above-mentioned multi-level insurance recommendation result includes a first-level recommendation result. The first-level recommendation result is at least one recommended insurance type, and the second-level recommendation result includes at least one recommended insurance product type of the above-mentioned recommended insurance type.

To further illustrate, the above-mentioned first-level recommendation result includes at least one recommended insurance type, such as accident insurance, medical insurance, cancer insurance, disability insurance, life insurance, major injury and illness insurance, and the like. The above-mentioned second-level recommendation result includes at least one recommended insurance product type of the above-mentioned recommended insurance types. Taking accident insurance as an example, accident insurance with medical expenses or accident insurance with subsidies for hospitalization, etc. Then, it can further include a third-level recommendation result, where the third-level recommendation result is the amount of at least one claim payment item of the recommended insurance product type in the above-mentioned recommended insurance types, for example, the accidental death insurance benefit is 5 million, or the hospitalization day A subsidy of 3,000 yuan, etc. In addition, it further includes the recommended insurance products of the fourth-level recommendation results. The above-mentioned fourth-level recommendation results are generated according to the above-mentioned types of insurance, the above-mentioned insurance product types and the above-mentioned claims and payment items. For example, an accident of an insurance company is generated. Insurance plans, etc. The above is only an example, not a limitation.

In another embodiment, the above-mentioned different levels can be arbitrarily combined and matched to generate the above-mentioned multi-level insurance recommendation result, for example, a multi-level insurance recommendation result composed of a first-level insurance type and a fourth-level insurance product, Or a multi-level recommendation result consisting of the first-level insurance types, the second-level recommended insurance product types, and the fourth-level recommended insurance products.

Furthermore, the above-mentioned personalized recommendation device 300 for insurance products further includes a user behavior collection device for collecting user behavior through the user device, and the above-mentioned user behavior history information is determined by the above-mentioned user behavior. In an actual implementation, when the user uses the user device 100 to click, browse or select information such as insurance conditions, insurance products, insurance types, insurance types, etc., these clicks, browsing or selection actions will be transmitted to the user device 100 The user behavior collection device 600 connected to the signal is stored in the user historical behavior database 204 after proper data processing and conversion. In other words, the user historical behavior information stored in the user historical behavior database 400 It is determined by user behavior. Moreover, each user has a unique identification and is associated with the user's historical behavior information.

A plurality of models to be trained are provided in the training device 320 , and the training device 320 will, according to the user historical behavior information stored in the user historical behavior database 400 and the insurance product information stored in an insurance product database 500 , , and train a plurality of models to be trained to obtain a plurality of trained models and a plurality of trained data, and store them in the data and model database 330 for generating at least one person when a recommendation algorithm requirement occurs. Multi-level insurance recommendation results.

The above-mentioned input information includes a user information, a device information and the above-mentioned user historical behavior information. The above-mentioned device information is the type of the user's device, such as a mobile phone, a tablet, or a computer. The above-mentioned plurality of user information includes at least one of gender, age, financial status, work type, family composition, transportation habits, living habits, leisure interests, and insurance specific needs. Next, the front-end device 200 sends the above at least one insurance recommendation result to the user device 100 and provides it to the user.

To further illustrate, generate personalized insurance recommendation results by entering information. Based on the above-mentioned plurality of user information, the basic insurance needs of the user can be understood, and the potential needs of the user can be predicted based on the above-mentioned user historical behavior information, so that the operating device 310 can generate at least one personalized multi-level based on the above. Insurance recommendation results, and the insurance recommendation results can be closer to the user's own needs.

For example, the user inputs personal information through the user device 100, such as a male, age 35, married with a son, occupation as a deliveryman, etc. The above-mentioned personalized recommendation device 300 for insurance products is based on the above-mentioned information. After obtaining the most suitable data and model from the data and model database 330, at least one personalized multi-level insurance recommendation result is generated, and one or more suitable insurance types, such as accident insurance, Medical insurance, cancer insurance, disability insurance, life insurance, and major injury and illness insurance (the first-level recommendation results are for example only, not limited), etc. Furthermore, recommend one or more of each insurance type It is assumed that the user chooses the option of accident insurance due to occupational reasons. The insurance product types include basic accident insurance, accident insurance with medical expenses, accident insurance with hospitalization allowance, and field workers often need to add insurance. The traffic accident insurance and the accident insurance for frequent travel (the second-level recommendation results are only for example, not limited by this). Next, recommend one or more payment item amounts for the type of insurance product, such as accident insurance death benefit of 5 million yuan, hospitalization day amount of 3,000 yuan, and actual hospitalization payment of 100,000 yuan (the third-level recommendation results, only For example, not limited), etc., and recommend products according to the user's personal information in each insurance product type.

According to the above example, users can choose the option of accident insurance according to their own occupational reasons (the first-level recommendation results are for example only, not limited), so as to understand the content of various accident insurances, such as the frequent occurrence of field workers. The traffic accident insurance that needs to be insured, and the user has a child, may accompany the child on trips often, and can also learn about the accident insurance for frequent trips (the second-level recommendation results are for example only, not limited). In addition, the personalized recommendation device 300 for insurance products will also directly recommend one or more insurance products of a certain insurance company according to the needs of the user (the fourth-level recommendation result is for illustration only and not limited thereto), Users can choose according to their own financial ability and preferences at that time.

In addition, through the user device 100, the user selects the type of insurance, the type of insurance product, adjusts the amount of the benefit item and the insurance product in the at least one personalized multi-level insurance recommendation result, and cooperates with historical user behavior information, such as the user's choice of Information such as insurance conditions, insurance types, insurance types, payment items, insurance companies, insurance specifications and attributes, adjustment of payment limits, consultation and message, etc., the above information will be transmitted to the user's collection device, and after appropriate data processing and conversion, it will be stored for use. The user history behavior database 400 is used for continuously adjusting the above-mentioned training model.

To sum up, the novel insurance product personalization recommendation device 300 can provide users with a multi-level insurance recommendation result according to their own conditions and needs, and each insurance product type has its recommended product ranking. It allows users to understand the relationship between the order of their recommended results and consumer needs, so that users can choose the most suitable insurance products for their needs. In addition, the user's historical behavior data will be stored in the user's historical behavior database for the purpose of continuously adjusting the above-mentioned training model, so as to strengthen the needs of the new type of insurance product personalization recommendation device for users accuracy.

The present invention is only disclosed in preferred embodiments herein. However, any person skilled in the art should understand that the above-mentioned embodiments are only used to describe the present invention and are not intended to limit the scope of the claimed patent rights of the present invention. All changes or substitutions that are equal or equivalent to the above embodiments should be construed as being covered within the spirit or scope of the present invention. Therefore, the protection scope of this new model shall be defined by the following patent application scope.

100: User device 200: Front end device 300: Personalized recommendation device for insurance products 310: Running Device 320: Training Device 330: Data and Model Database 400: User history behavior database 500: Insurance product database 600: User behavior collection device

The accompanying drawings are provided so that those skilled in the art to which the present invention pertains may further understand the present invention, and are incorporated in and constitute a part of the description of the present invention. The drawings illustrate exemplary embodiments of the invention, and together with the description of the invention serve to explain the principles of the invention. The following is a brief description of each figure of this new model: FIG. 1 is a functional block diagram of a personalization recommendation device for insurance products according to an embodiment of the present invention.

100: User device

200: Front end device

300: Personalized recommendation device for insurance products

310: Running Device

320: Training Device

320: Data and Model Database

400: User history behavior database

500: Insurance product database

600: User behavior collection device

Claims (10)

A personalized recommendation device for insurance products, which outputs at least one multi-level insurance recommendation result according to a received recommendation algorithm requirement and an input information, wherein the personalized recommendation device for insurance products includes: A training device is provided with a plurality of modules to be trained, and is used to train the modules to be trained according to a user's historical behavior information and an insurance product information, so as to obtain a plurality of trained models and a plurality of trained data , and stored in a data and model database; and a running device, connected to the training device by a signal, for selecting at least one training model and selecting at least one trained model from the trained modules and the trained data stored in the data and model database according to the requirements of the recommendation algorithm training data, and running the selected at least one trained module based on the selected at least one trained data and the input information, thereby generating the multi-level insurance recommendation result; Wherein, the multi-level insurance recommendation result includes a first-level recommendation result and an X-level recommendation result, where X is greater than or equal to 2. The device for personalized recommendation of insurance products according to claim 1, wherein the multi-level insurance recommendation result further includes a J-th level recommendation result, wherein J is not equal to X. The device for personalizing insurance products according to claim 1, wherein the first-level recommendation result includes at least one recommended insurance type, the second-level recommendation result includes at least one recommended insurance product type of the recommended insurance type, and the third-level recommendation result includes at least one recommended insurance product type. The hierarchical recommendation result is the amount of at least one claim payment item of the recommended insurance product type of the recommended insurance type. The device for personalizing insurance product recommendation according to claim 1, wherein the multi-level recommendation result further includes a fourth-level recommendation result, and the fourth-level recommendation result is at least one recommended insurance product. The personalized recommendation device for insurance products as described in claim 1, further comprising: A user behavior collection device is used to collect a user behavior through a user device, and the user behavior history information is determined by the user behavior. The personalized recommendation device for insurance products as claimed in claim 1, wherein the users have a unique identification and are associated with the user's historical behavior information. The personalized recommendation device for insurance products as described in claim 1, further comprising: A front-end device is connected to the running device by a signal, and is used for receiving the recommendation algorithm requirement and the input information. The personalized recommendation device for insurance products as described in claim 1, further comprising: The data and model database, the signal connects the running device and the training device. The personalized recommendation device for insurance products as described in claim 1, further comprising: a user history behavior database, the signal is connected to the training device for storing the user history information, and transmitting the user behavior history information to the training device; and An insurance commodity database, the signal is connected to the training device, and is used for storing a plurality of insurance commodity information. The personalized recommendation device for insurance products according to claim 1, wherein the input information includes a user information, a device information and the user's historical behavior information.

Images