Insurance product recommendation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an insurance product recommendation method and device, electronic equipment and a storage medium, belongs to the technical field of artificial intelligence, and is applied to the field of financial science and technology. The method comprises the steps of performing portrait construction on a target object based on original insurance data to obtain original portrait data, and performing collaborative filtering and screening based on an original user tag and original insurance product features to obtain candidate objects similar to the target object, reference insurance product data of the candidate objects are obtained, a candidate insurance product list is screened out from the reference insurance data, the candidate insurance product list is a list of unpurchased insurance of the target object and serves as a list possibly interested by the target object, collaborative filtering and screening are carried out on the basis of original insurance product features, and a supplementary insurance product list is obtained; therefore, insurance product recommendation is carried out on the target object according to the candidate insurance product list and the supplementary insurance product list, and the insurance product recommendation efficiency and accuracy are improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology and is applied in financial technology scenarios, and in particular to an insurance product recommendation method and device, electronic device, and storage medium. Background Art

[0002] Intelligent recommendations can be used to mine data to recommend content of interest to users, such as items, articles, videos, and music. This can improve user efficiency in selecting the content they need, enhancing the user experience. For example, in the financial sector, users often face information asymmetry and difficulty choosing insurance products. Users may not be clear about their risk profile and true insurance needs, making it difficult to select the right insurance product from a wide range of options. Therefore, how to use intelligent recommendation technology to recommend the insurance products that users need has become a pressing issue. Summary of the Invention

[0003] The main purpose of the embodiments of the present application is to propose an insurance product recommendation method and device, electronic device, and storage medium, aiming to improve the efficiency and accuracy of insurance product recommendations.

[0004] To achieve the above objectives, a first aspect of an embodiment of the present application provides an insurance product recommendation method, the method comprising:

[0005] Obtain the original insurance data of the target object;

[0006] Constructing a profile of the target object based on the original insurance data to obtain original profile data of the target object; wherein the original profile data includes original user tags and original insurance product features, and the original user tags are multiple tags;

[0007] Perform collaborative filtering based on the original user tags and the original insurance product features to obtain candidate objects;

[0008] Obtain reference insurance product data of the candidate subject, and filter out a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is data of insurance products purchased by the candidate subject, and the list of candidate insurance products is a list of insurance products not purchased by the target subject;

[0009] Perform collaborative filtering based on the characteristics of the original insurance products to obtain a list of supplementary insurance products;

[0010] Recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

[0011] In some embodiments, performing collaborative filtering based on the original insurance product features to obtain a list of supplementary insurance products includes:

[0012] Obtaining preset insurance from a preset insurance resource pool, and performing feature extraction on the preset insurance to obtain preset insurance features;

[0013] Calculating similarity between the original insurance product features and the preset insurance features to obtain insurance product similarity between the original insurance product features and the preset insurance features;

[0014] The supplementary insurance product list is selected from the insurance resource screening pool based on the insurance product similarity.

[0015] In some embodiments, recommending insurance products to the target subject based on the candidate insurance product list and the supplementary insurance product list includes:

[0016] Screening a first selected insurance product from the candidate insurance product list; wherein the first selected insurance product has a first weight;

[0017] Filtering a second selected insurance product from the list of supplementary insurance products; wherein the second selected insurance product has a second weight;

[0018] Performing weighted calculation on the first selected insurance product and the second selected insurance product based on the first weight and the second weight to obtain a list of selected insurance products;

[0019] Filtering target insurance products from the selected insurance product list;

[0020] Recommend insurance products to the target object based on the target insurance product.

[0021] In some embodiments, after recommending an insurance product to the target subject based on the target insurance product, the method further includes: pushing a recommendation report, specifically including:

[0022] Obtaining product information of the target insurance product to obtain target product information;

[0023] generating an insurance recommendation report based on the target insurance information and the characteristics of the original insurance product;

[0024] Push the insurance recommendation report to the target object.

[0025] In some embodiments, the original portrait data further includes clusters, and constructing a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object includes:

[0026] Extracting key information from the original insurance data to obtain original key information;

[0027] Performing feature encoding on the original key information to obtain preliminary encoding features;

[0028] The target object is clustered based on the preliminary coding features to obtain the cluster to which the target object belongs, the original user label, and the original insurance product features.

[0029] In some embodiments, after constructing a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, the method further includes:

[0030] Configuring display colors for the clusters according to a preset mapping configuration table;

[0031] Configuring a display identifier for the original user tag and the original insurance product feature according to the mapping configuration table;

[0032] The original portrait data is visually displayed according to the display color and the display identifier.

[0033] In some embodiments, after constructing a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, the method further includes: updating the data, specifically including:

[0034] In response to a periodic update request, collecting new insurance data at a preset period; wherein the new insurance data includes new insurance products;

[0035] updating the insurance resource pool according to the new insurance data; wherein the new insurance product is used as the preset insurance;

[0036] The original portrait data is updated based on the updated insurance resource pool.

[0037] To achieve the above-mentioned objectives, a second aspect of an embodiment of the present application provides an insurance product recommendation device, comprising:

[0038] Insurance data acquisition module, used to obtain the original insurance data of the target object;

[0039] A user portrait construction module is used to construct a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object; wherein the original portrait data includes the original user label and the original insurance product feature, and the original user label is a multi-label;

[0040] A first collaborative filtering module is configured to perform collaborative filtering based on the original user tags and the original insurance product features to obtain candidate objects;

[0041] An insurance product screening module, configured to obtain reference insurance product data of the candidate subject and screen a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is data of insurance products purchased by the candidate subject, and the list of candidate insurance products is a list of insurance products not purchased by the target subject;

[0042] A second collaborative filtering module is configured to perform collaborative filtering based on the characteristics of the original insurance products to obtain a list of supplementary insurance products;

[0043] An insurance product recommendation module is used to recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

[0044] To achieve the above-mentioned purpose, the third aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the method described in the first aspect when executing the computer program.

[0045] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a storage medium, which is a computer-readable storage medium and stores a computer program. When the computer program is executed by a processor, the method described in the first aspect is implemented.

[0046] The insurance product recommendation method and device, electronic device, and storage medium proposed in the embodiments of the present application can construct a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, which includes the original user label and the original insurance product characteristics, and perform collaborative filtering based on the original user label and the original insurance product characteristics to obtain candidate objects similar to the target object, obtain reference insurance product data of the candidate objects, and filter out a list of candidate insurance products from the reference insurance data. The list of candidate insurance products is a list of insurance that the target object has not purchased and serves as a first list that the target object may be interested in. Collaborative filtering is performed based on the original insurance product characteristics to obtain a supplementary insurance product list as a second list that the target object may be interested in, thereby recommending insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list, thereby improving the efficiency and accuracy of insurance product recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of the insurance product recommendation method provided in an embodiment of the present application;

[0048] Figure 2 yes Figure 1 Flowchart of step 102 in FIG.

[0049] Figure 3 is another flow chart of the insurance product recommendation method provided in an embodiment of the present application;

[0050] Figure 4 yes Figure 1 Flowchart of step 105 in FIG.

[0051] Figure 5 yes Figure 1 Flowchart of step 106 in FIG.

[0052] Figure 6 This is another flow chart of the insurance product recommendation method provided in the embodiment of the present application;

[0053] Figure 7 This is a schematic diagram of the structure of the insurance product recommendation device provided in an embodiment of the present application;

[0054] Figure 8 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0056] It should be noted that although the device schematics illustrate functional module divisions and the flowcharts illustrate logical sequences, in certain circumstances, the steps shown or described may be performed in a sequence that differs from the module divisions in the device or the sequence in the flowcharts. The terms "first," "second," and so on, in the specification, claims, and drawings, are used to distinguish similar items and are not necessarily used to describe a specific sequence or precedence.

[0057] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0058] First, let’s analyze some of the terms used in this application:

[0059] Artificial intelligence (AI) is a new technical discipline that studies and develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. A branch of computer science, AI seeks to understand the essence of intelligence and produce new intelligent machines that can respond in a manner similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thinking. It also encompasses the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, to perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results.

[0060] Natural language processing (NLP): NLP uses computers to process, understand, and apply human languages ​​(such as Chinese and English). A branch of artificial intelligence, NLP is an interdisciplinary field between computer science and linguistics, often referred to as computational linguistics. Natural language processing encompasses grammatical analysis, semantic analysis, and discourse comprehension. Natural language processing is commonly used in technical fields such as machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, information intent recognition, information extraction and filtering, text classification and clustering, public opinion analysis, and opinion mining. It encompasses data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research related to language processing, and linguistics research related to language computing.

[0061] Intelligent recommendation can be based on data mining to recommend content of interest to users, such as recommended items, articles, videos, music, etc. Through intelligent recommendation, the efficiency of users selecting the content they need can be improved, and the user experience can be enhanced. The implementation principle of intelligent recommendation is usually based on analyzing data such as users' historical behavior, personal preferences, current context information, etc. to predict the content that users may be interested in. The implementation methods of intelligent recommendation mainly include the following five aspects: (1) Content-Based Filtering: This method is mainly based on the attributes and characteristics of the item itself. By analyzing the user's past behavior, the user's preference patterns are identified, and items with similar attributes are recommended. (2) Model-Based Methods: This method uses machine learning models to predict user preferences and is suitable for processing complex user behavior data. (3) Context-Aware Recommender Systems: This method considers the user's context information (such as time, location, etc.) to provide more personalized recommendations. (4) Collaborative Filtering: This includes user collaborative filtering and item collaborative filtering. User collaborative filtering is primarily based on the similarity between users, recommending items by finding other users with similar preferences to the target user; item collaborative filtering focuses on the similarity between items, recommending other items with similar characteristics to items that the user has already liked. (5) Hybrid Recommender Systems: This combines multiple recommendation methods to improve the accuracy and diversity of recommendations.

[0062] Taking the financial sector as an example, users often face information asymmetry and difficulty choosing insurance products. Users may not be clear about their risk profile and true insurance needs, making it difficult to filter out suitable products from a wide range of options. Furthermore, current intelligent recommendation methods often rely on basic user information and limited purchase history, failing to fully tap into customers' potential needs and behavioral patterns, resulting in inaccurate recommendations.

[0063] Based on this, the embodiments of the present application provide an insurance product recommendation method and device, electronic device, and storage medium, aiming to improve the efficiency and accuracy of insurance product recommendations.

[0064] The insurance product recommendation method and device, electronic device, and storage medium provided in the embodiments of the present application are specifically illustrated through the following embodiments. First, the insurance product recommendation method in the embodiments of the present application is described.

[0065] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to achieve optimal results.

[0066] Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0067] The insurance product recommendation method provided in the embodiment of the present application relates to the field of artificial intelligence technology. The insurance product recommendation method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or as a server cluster or distributed system composed of multiple physical servers, or as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the insurance product recommendation method, etc., but is not limited to the above forms.

[0068] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments in which tasks are performed by remote processing devices connected via a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0069] It should be noted that in each specific embodiment of the present application, when it comes to the need to perform relevant processing based on data related to the user's identity or characteristics, such as user personal information, user attribute information, user historical data, etc., the user's permission or consent will be obtained first, and the collection, use and processing of such data will comply with relevant laws, regulations and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0070] Figure 1 This is an optional flowchart of the insurance product recommendation method provided in the embodiment of the present application. Figure 1 The method may include but is not limited to steps 101 to 106.

[0071] Step 101: Obtain original insurance data of the target object;

[0072] Step 102: construct a profile of the target object based on the original insurance data to obtain the original profile data of the target object; wherein the original profile data includes the original user label and the original insurance product features, and the original user label is a multi-label.

[0073] Step 103: Perform collaborative filtering based on the original user tags and original insurance product features to obtain candidate objects;

[0074] Step 104: Obtain reference insurance product data for the candidate subject, and filter out a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is data on insurance products purchased by the candidate subject, and the list of candidate insurance products is a list of insurance products not purchased by the target subject;

[0075] Step 105: Perform collaborative filtering based on the features of the original insurance products to obtain a list of supplementary insurance products;

[0076] Step 106: Recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

[0077] Steps 101 to 106 shown in the embodiment of the present application can construct a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, which includes the original user label and the original insurance product characteristics, and perform collaborative filtering based on the original user label and the original insurance product characteristics to obtain candidate objects similar to the target object, obtain reference insurance product data of the candidate objects, and filter out a list of candidate insurance products from the reference insurance data. The list of candidate insurance products is a list of insurance that the target object has not purchased and serves as a first list that the target object may be interested in. Collaborative filtering is performed based on the original insurance product characteristics to obtain a supplementary insurance product list as a second list that the target object may be interested in, thereby recommending insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list, thereby improving the efficiency and accuracy of insurance product recommendations.

[0078] In some embodiments, step 101 may include, but is not limited to, data collection and data preprocessing, specifically including:

[0079] Obtain the target object's historical purchase data; wherein, the historical purchase data refers to the insurance data purchased by the target object, including policyholder information, guarantor information, and insurance-related information; policyholder information includes the policyholder's name, ID number, gender, etc., guarantor information includes the insured's name, ID number, etc., and insurance-related information includes insurance product name, coverage, premium, insured amount, policy term, claims records, and other information.

[0080] Preprocess the historical purchase data to obtain the original insurance data; specifically, it may include but is not limited to cleaning and standardizing the historical purchase data, specifically, it may include but is not limited to deleting duplicate data, processing missing values, and correcting outliers, etc., wherein processing missing values ​​may specifically include but is not limited to: using mean filling, median filling or model-based prediction filling methods, for example, for missing annual income, mean filling or median filling may be used, which is not limited in the embodiments of the present application; correcting outliers, specifically, it may include but is not limited to: using statistical methods or setting thresholds based on domain knowledge, which is not limited in the embodiments of the present application.

[0081] See also Figure 2 In some embodiments, step 102 may include, but is not limited to:

[0082] Step 201: extract key information from original insurance data to obtain original key information;

[0083] Step 202: feature encode the original key information to obtain preliminary encoding features;

[0084] Step 203 : clustering the target object based on the preliminary coding features to obtain the cluster to which the target object belongs, the original user label, and the original insurance product features.

[0085] In step 201 of some embodiments, key information is extracted from the original insurance data. The obtained original key information includes basic information of the insurance data purchased by the target object, insurance information, such as age, gender, income, occupation, family status, etc. The types of original key information may include, but are not limited to, numerical, categorical, and textual. The original key information of numerical type may include, but is not limited to: premium, insured amount, deductible, insured age range (such as 18-60 years old), coverage period (years), compensation ratio, cash value, etc. The original key information of categorical type may include, but is not limited to: insurance type (critical illness insurance / medical insurance / accident insurance), coverage (cancer / cardiovascular / accidental disability), insured population (adults / children / elderly), payment method (annual payment / monthly payment). The original key information of text type may include, but is not limited to: clause description (such as "covering imported drugs"), exemption clause (such as "no compensation for congenital diseases"), etc.

[0086] In step 202 of some embodiments, the method of feature encoding the original key information may include but is not limited to: numerical encoding, label encoding, and the method of feature encoding the original key information may also include a method of embedding vectors, so that the numerical type, category type and text type can be merged into a unified vector representation. In one application scenario, the user's basic information can be numerically encoded, which may specifically include but is not limited to: converting "gender" to 0 and 1, where 0 represents male and 1 represents female; one-hot encoding the occupation (One-Hot Encoding), for example, the occupation includes white-collar, blue-collar, freelance, etc., where "white-collar" is one-hot encoded to obtain [1,0,0] and "blue-collar" is one-hot encoded to obtain [0,1,0]. In other application scenarios, label encoding is performed for insurance information, such as insurance type (such as life insurance, health insurance, etc.), and numerical features such as the insurance amount and policy term are normalized so that the features of the insurance products are in the same numerical range. The method of feature encoding the original key information in the embodiment of this application is not limited.

[0087] Through the above embodiment, the features can be standardized (for example, the mean is 0 and the variance is 1) to avoid the dimensionality difference (for example, different units of age and insurance amount) affecting the dimensionality reduction result.

[0088] In step 203 of some embodiments, user feature data can be clustered based on the K-Means clustering algorithm. Among them, the K-Means clustering algorithm is an unsupervised learning algorithm that divides the data set into K clusters so that the data points within the cluster are as similar as possible, while the data points between different clusters are as different as possible; the K-Means clustering algorithm can continuously optimize the center position of the cluster in an iterative manner until a certain termination condition is met. In the clustering process of the embodiment of the present application, a suitable distance measurement method (such as Euclidean distance, Manhattan distance, etc.) is selected to measure the similarity between users. By continuously adjusting the clustering parameters, such as the number of clusters K, a more reasonable user classification result is obtained to form a preliminary customer portrait. In other embodiments, the target object can also be clustered based on a hierarchical clustering algorithm, which is not limited in the embodiment of the present application.

[0089] During clustering in step 203, the center vector of each cluster is calculated as the typical characteristic of the object in that cluster. For example, objects over 60 years old are considered the elderly group. The original user label can be a user label with n dimensions or multiple labels. For example, the original user label can represent the target object's customer size, capital scale, industry risk, historical claims, etc. The similarity between each label of the target customer and similar customers is then calculated to facilitate collaborative filtering in the subsequent steps.

[0090] The technical principles of the above-mentioned portrait construction are illustrated using an application scenario. As shown in Table 1 below, Cluster 0 is a group of young people with an average age of 28. They are young, low-risk customers. The cluster characteristics are: young, low premiums, and no claims. They may be new employees taking out insurance for the first time. Cluster 1 is a group of middle-aged people with an average age of 45. They are high-net-worth customers. The cluster characteristics are: middle-aged, high premiums, and high insurance amounts. They may be business owners or high-income people. Cluster 2 is a group of older people with an average age of 60. They are high-risk elderly customers. The cluster characteristics are: advanced age, low insurance amounts, but a high number of claims. They may be older policyholders with higher risks.

[0091] Cluster ID age Premium (10,000 yuan / year) Insured amount (10,000 yuan) Profession area Number of claims 0 28 0.6 10 white collar First-tier cities 0 1 45 2 120 blue-collar workers Second-tier cities 1 2 61 1.2 60 retirees Third-tier cities 3

[0092] Table 1

[0093] See also Figure 3 In some embodiments, after step 102, the insurance product recommendation method further includes: visually displaying the original portrait data, which may specifically include but is not limited to:

[0094] Step 301, configuring display colors for clusters according to a preset mapping configuration table;

[0095] Step 302: configuring a display identifier for the original user tag and the original insurance product feature according to the mapping configuration table;

[0096] Step 303: Visually display the original portrait data according to the display color and display identifier.

[0097] In some embodiments, dimensionality reduction techniques such as principal component analysis (PCA) or t-SNE (t-Distributed Stochastic Neighbor Embedding) may be used to visualize the original portrait data, which is not limited in the embodiments of the present application. PCA is a linear transformation used to transform data into a new coordinate system so that the first largest variance of any data projection is on the first coordinate (called the first principal component), the second largest variance is on the second coordinate (the second principal component), and so on; PCA is often used to reduce the dimensionality of a data set. t-SNE is a nonlinear algorithm for dimensionality reduction and data visualization, suitable for processing high-dimensional data.

[0098] In some embodiments, before executing step 301, the insurance product recommendation method further includes: pre-configuring a mapping configuration table, wherein the mapping configuration table is configured with a mapping relationship between clusters and display colors, and a mapping relationship between original user labels and display identifiers of original insurance product features. The display identifiers can be shaped like arrows, triangles, or other symbols. In one application scenario, for example, the display color is set to blue to represent the cluster "young new customers," and the display color is set to red to represent the cluster "high-risk customers." The display identifier is set to a triangle to represent the original user label "high-risk customers," and the display identifier is set to a five-pointed star to represent the original insurance product feature "children's critical illness insurance." The pre-configured mapping configuration table facilitates querying the relevant display colors and display identifiers in steps 301 and 302, thereby configuring display colors for clusters and configuring display identifiers for original user labels and original insurance product features. In other embodiments, the display identifier colors can also be configured based on the original user label and original insurance product feature. For example, the original insurance product feature "children's critical illness insurance" is represented by a blue five-pointed star, and the original user label "high-risk customer" is represented by a red triangle.

[0099] In step 303 of some embodiments, the original portrait data is visualized based on the configured display colors and display identifiers, so as to facilitate an intuitive understanding of the distribution and classification of customers, and further facilitate the formulation of precision marketing or risk management strategies.

[0100] In some embodiments, step 103 may employ a user-based collaborative filtering method to screen out candidate objects, which are groups similar to the target object. This collaborative filtering method can be used to find other customer groups that are close to the target object in feature space. In some application scenarios, the Pearson Correlation Coefficient or cosine similarity can be used to measure the similarity between users. This embodiment of the present application is not limited to this.

[0101] After candidate objects similar to the target object are screened out according to step 103, insurance products (reference insurance data) purchased by similar customer groups (candidate objects) but not purchased by the target object are obtained according to step 104, thereby obtaining a list of candidate insurance products. Specifically, in one application scenario, a customer group similar to target object A includes candidate objects B and candidate objects C. Target object A has purchased insurance product p1. Reference insurance product data purchased by candidate object B includes insurance products p1, insurance product p2, insurance product q1, insurance product q2, and insurance product q3. Reference insurance product data purchased by candidate object C includes insurance products p2, insurance product p3, insurance product q1, and insurance product q4. Insurance products p2, insurance product p3, insurance product q1, insurance product q2, insurance product q3, and insurance product q4 are then used as the list of candidate insurance products.

[0102] See also Figure 4 In some embodiments, step 105 may include, but is not limited to:

[0103] Step 401: Obtain preset insurance from a preset insurance resource pool, and perform feature extraction on the preset insurance to obtain preset insurance features;

[0104] Step 402: Calculate the similarity between the original insurance product features and the preset insurance features to obtain the insurance product similarity between the original insurance product features and the preset insurance features;

[0105] Step 403: Select a supplementary insurance product list from the insurance resource screening pool based on the insurance product similarity.

[0106] In some embodiments, step 105 may implement screening of insurance products based on an item-based collaborative filtering method to obtain a list of supplementary insurance products that can fill the protection gap of the insurance products already purchased by the target object.

[0107] Prior to executing step 401, the insurance product recommendation method further includes: constructing an insurance resource pool, which may specifically include, but is not limited to: obtaining insurance data of the insured to obtain preset insurance data, wherein the preset insurance data includes preset insurances for multiple insured persons; and pre-processing the preset insurance data. The preset insurance data refers to insurance data purchased by the insured, including information about the policyholder, the guarantor, and insurance-related information. The policyholder information includes the policyholder's name, ID number, and gender; the guarantor information includes the insured's name and ID number; and the insurance-related information includes information such as the insurance product name, coverage, premium, insured amount, policy term, and claims settlement history. Pre-processing the preset insurance data may specifically include but is not limited to cleaning and standardizing the preset insurance data, including but not limited to deleting duplicate data, processing missing values, and correcting outliers, among which processing missing values ​​may specifically include but is not limited to: using mean filling, median filling, or model-based prediction filling methods, for example, for missing annual income, mean filling or median filling may be used, which is not limited in this embodiment of the present application; correcting outliers may specifically include but is not limited to: setting thresholds through statistical methods or based on domain knowledge, which is not limited in this embodiment of the present application. The way in which the preset insurance data is pre-processed in this embodiment of the present application is basically the same as the way in which the historical purchase data is pre-processed in step 101 above.

[0108] In some embodiments, the preset insurance features obtained through step 401 may include age, gender, income, occupation, family status, etc. used to represent basic information features of the user, and may also include features used to represent insurance product features, such as: premium, insured amount, deductible, insured age range (such as 18-60 years old), coverage period (years), compensation ratio, cash value, insurance type (critical illness insurance / medical insurance / accident insurance), coverage (cancer / cardiovascular and cerebrovascular / accidental disability), insured population (adults / children / elderly), payment method (annual payment / monthly payment), etc. In one application scenario, the preset insurance in the insurance resource pool may be, for example, "M Company Children's Critical Illness Insurance", and the preset insurance features obtained by feature extraction of the preset insurance may include: (1) Insurance type: critical illness insurance; (2) Insured population: children (0-18 years old); (3) Coverage: leukemia, cancer; (4) Premium: 3,000 yuan / year.

[0109] In step 402 of some embodiments, the similarity between different insurance products can be calculated using methods such as cosine similarity or Euclidean distance. For example, for cosine similarity, the degree of similarity between the feature vectors of two insurance products is measured by calculating the cosine value of the angle between them; for Euclidean distance, the distance between the two feature vectors in space is calculated, and the smaller the distance, the higher the similarity. Cosine similarity is used in scenarios where high-dimensional sparse features (such as text, one-hot encoding, etc.) are used. Euclidean distance is used in scenarios where low-dimensional dense numerical features (such as premiums, insured amounts, etc.) are used. The insurance product similarity calculated in step 402 can be used in step 403 to screen the insurance resource pool, thereby selecting a list of supplementary insurance products. The list of supplementary insurance products can be used to fill possible protection gaps, for example, to fill protection gaps such as low insured amounts of purchased health insurance, non-coverage of certain major diseases, and limited coverage of accident insurance.

[0110] Insurance product name Similarity Matching instructions Children's Medical Insurance A1 0.86 Similar population, covering similar disease types Children's critical illness insurance B1 0.78 Similar premiums and overlapping coverage School insurance C1 0.66 Additional accidental medical liability

[0111] Table 2

[0112] In one application scenario, as shown in Table 2 above, the list of supplementary insurance products with similar characteristics to the original insurance product "M Company's Children's Critical Illness Insurance" includes three insurance products: Children's Medical Insurance a1, Children's Critical Illness Insurance b1, and School Insurance c1. Among them, the insurance product "Children's Medical Insurance a1" has the highest similarity, which is 0.86; the insurance product "Children's Critical Illness Insurance b1" has a similarity of 0.86.

[0113] In a specific scenario, as shown in Table 2 above, an insurance product similar to "M Company's Children's Critical Illness Insurance" is recommended to target object A. The characteristics of this insurance product include: (1) Insurance type: critical illness insurance; (2) Insured population: children (0-18 years old); (3) Coverage: leukemia, cancer; (4) Premium: 3,000 yuan / year. Similar products are calculated for insurance products: cosine similarity is used to calculate text descriptions (such as "covering leukemia" vs. "covering cancer"). Combined with numerical features (premium difference weights reduce similarity), the three recommended insurance products shown in Table 2 are obtained. Among them, the insurance product "Children's Medical Insurance a1" has the highest similarity of 0.86; the insurance product "Children's Critical Illness Insurance b1" has a similarity of 0.86.

[0114] In some other embodiments, if data on new insurance products is lacking, content similarity (e.g., clause text) or average similarity based on similar products can be used. Furthermore, embodiments of the present application periodically recalculate the similarity matrix (e.g., monthly) to accommodate product iterations and display matching features in recommendation results (e.g., "Recommended due to cancer coverage") to enhance user trust.

[0115] In some embodiments, after step 102, the insurance product recommendation method further includes: updating data, which may specifically include but is not limited to:

[0116] In response to a periodic update request, collecting new insurance data at a preset period; wherein the new insurance data includes new insurance products;

[0117] Update the insurance resource pool based on new insurance data; in particular, use the new insurance products as default insurance;

[0118] Update the original portrait data based on the updated insurance resource pool.

[0119] In some embodiments, a scheduled task can be suggested to obtain a periodic update request, thereby collecting new insurance data in a preset period, which can be set to weekly, monthly, quarterly, etc. New insurance products refer to products that have not appeared on the market and new products released by insurance companies; in addition to new insurance products, new insurance data can also include new data of historical customers, such as new claims data, new insurance data, new customers, etc. of historical customers. Therefore, the insurance resource pool can be updated according to the new insurance data, and the new insurance product can be used as the preset insurance, and the new customer can be used as the historical customer in the insurance resource pool. Based on the updated insurance resource pool, the original portrait data is updated, thereby updating the original user label, original insurance product characteristics, clustering clusters and other information of the target object, thereby changing the real-time update and dynamic adjustment of the corresponding data machine to adapt to market changes and changes in customer needs.

[0120] In an embodiment of the present application, by establishing a scheduled task, the latest insurance data of the customer and the market insurance product information are re-collected and processed at a preset period (such as weekly or monthly) to update the customer portrait and the insurance product characteristics of the insurance resource pool, so that when new insurance products appear on the market or the terms of existing insurance products change, the insurance resource pool and the above-mentioned similarity calculation results can be updated in time.

[0121] See also Figure 5 In some embodiments, step 106 may include, but is not limited to:

[0122] Step 501: Screen a first selected insurance product from a list of candidate insurance products; wherein the first selected insurance product has a first weight;

[0123] Step 502: Filter a second selected insurance product from the list of supplementary insurance products; wherein the second selected insurance product has a second weight;

[0124] Step 503: Perform weighted fusion on the first selected insurance product and the second selected insurance product based on the first weight and the second weight to obtain a list of selected insurance products;

[0125] Step 504: Filter the target insurance product from the selected insurance product list;

[0126] Step 505: Recommend insurance products to the target object based on the target insurance product.

[0127] In some embodiments, steps 501 and 502 may combine the results of user-based collaborative filtering and item-based collaborative filtering to assign weights, for example, a first weight of 0.6 and a second weight of 0.4. In other embodiments, the first and second weights may be dynamically adjusted based on customer feedback and purchasing behavior to improve recommendation accuracy and satisfaction, for example, a first weight of 0.4 and a second weight of 0.6. For example, if a customer frequently purchases or expresses satisfaction with a recommended product, the weight of the corresponding recommendation method may be increased; if the customer is not interested in the recommendation result, the corresponding weight may be decreased. Compared to conventional techniques, the embodiments of the present application fully consider customer feedback on recommended products and promptly optimize the recommendation strategy based on actual customer response. k1 first selected insurance products are screened from the candidate insurance product list, and k2 second selected insurance products are screened from the supplementary insurance product list. The values ​​of k1 and k2 may be the same or different, for example, k1 may be 2 and k2 may be 3, or k1 may be 3 and k2 may be 3.

[0128] In step 503 of some embodiments, the results of user-based collaborative filtering and item-based collaborative filtering are comprehensively considered, and a weighted calculation is performed on the first selected insurance product and the second selected insurance product, using a first weight and a second weight of 0.4. The calculation result may be a score of the insurance product, which is then sorted based on the score to obtain a list of selected insurance products. In step 504, the insurance products in the list of selected insurance products are deduplicated and ranked in order, and the top N insurance products are selected as target insurance products, where N is a positive integer that can be greater than or equal to 2. The value of N can be determined based on actual needs, for example, N can be 3. Consequently, in step 505, the top N target insurance products are recommended to the target object.

[0129] The embodiment of the present application integrates two recommendation algorithms (user-based collaborative filtering and project-based coordinated filtering) and reasonably allocates weights to give full play to the respective advantages of the two recommendation algorithms and improve the comprehensiveness and accuracy of the recommendations.

[0130] In other application scenarios, for new customers, the embodiments of the present application can provide new customers with initial recommendations with certain reference value by analyzing the basic information and initial needs of the new customers, combining the general risk patterns and popular insurance products in the market. As the relevant data of the new customers accumulates, the recommendations will become more and more accurate.

[0131] See also Figure 6 In some embodiments, after step 505, the insurance product recommendation method further includes: generating a recommendation report, specifically including:

[0132] Step 601: Acquire product information of a target insurance product to obtain target product information;

[0133] Step 602: Generate an insurance recommendation report based on the target insurance information and the original insurance product characteristics;

[0134] Step 603: Push the insurance recommendation report to the target object.

[0135] In step 601 of some embodiments, the target product information is relevant information of the target insurance product, and may include, but is not limited to, the name of the target insurance product, insurance type, coverage, product parameters, insurance conditions, coverage period, relevant clauses, claims settlement and service procedures, etc., wherein the product parameters include the insured amount, premium, deductible, compensation ratio, etc., and the relevant clauses include optional liability clauses, exemption clauses, etc. As shown in Table 3 below, taking "M Company Children's Critical Illness Insurance" as an example, the target product information of "M Company Children's Critical Illness Insurance" includes the following:

[0136]

[0137] Table 3

[0138] In step 602 of some embodiments, the insurance recommendation report generated by combining the original insurance product characteristics and the target insurance information may include, but is not limited to, the name, features, advantages, applicable scenarios, and extended warranty of the target insurance product. Thus, in step 603, the insurance recommendation report is pushed to the target subject, allowing the target subject to select an insurance product of interest based on the insurance recommendation report. In embodiments of the present application, an insurance recommendation report can be generated through steps 601 to 603 and pushed to the target subject, allowing the target subject to clearly and intuitively understand the coverage information and coverage gaps of the insurance they have purchased, so that the target subject can select other insurance products as needed.

[0139] In one application scenario, the target object has purchased critical illness insurance (for example, a critical illness insurance with a coverage of 500,000 yuan and covering cancer). The insurance products that meet the target object's demand gap include medical insurance and accident insurance. The original insurance product characteristics of the target object are the product characteristics of the purchased critical illness insurance, and the target insurance information is the product information of the medical insurance and the product information of the accident insurance. The obtained insurance recommendation report can be combined with the following Table 4: The insurance recommendation report shows [I. Overview of existing protection], specifically showing at least the following four aspects of information: purchased insurance products, protection type, coverage amount, and covered risks (for example, leukemia, malignant tumors); The insurance recommendation report The report displays [II. Target Insurance Products], which specifically include two types: Children's Medical Insurance a1 and Children's Critical Illness Insurance b1, and displays the [Coverage Responsibilities], [Unique Advantages], [Applicable Scenarios] and other information of Children's Medical Insurance a1 and Children's Critical Illness Insurance b1 respectively; the insurance recommendation report also displays [III. Reasons for Recommendation], and presents it from at least the following three aspects: [Needs Matching], [Cost-Effectiveness], and [Needs Matching]; in addition, the insurance recommendation report also displays [IV. Next Steps], and gives at least the following suggestions: ① Prioritize medical insurance (annual premium of approximately 500 yuan); ② Contact customer service to receive a free family policy review service.

[0140] In another application scenario, a car insurance product is recommended to the target object B, wherein the relevant information of the target object B includes: 30 years old, male, having purchased "compulsory traffic insurance" and "third-party liability insurance"; based on the above step 103, a similar user group of the target object B is found, and "vehicle liability insurance" (purchase rate 60%) is recommended; based on similar products of "third-party liability insurance", "no deductible insurance" is recommended (similarity 0.85); and based on the above steps 501 to 505, weighted fusion is performed: after the comprehensive score is sorted, "no deductible insurance" (weight 0.6) and "vehicle liability insurance" (weight 0.4) are recommended.

[0141]

[0142] Table 4

[0143] Through the insurance recommendation method of the embodiment of the present application, a portrait of the target object can be constructed based on the original insurance data to obtain the original portrait data of the target object, which includes the original user label and the original insurance product characteristics. Collaborative filtering and screening are performed based on the original user label and the original insurance product characteristics to obtain candidate objects similar to the target object, obtain reference insurance product data of the candidate objects, and filter out a list of candidate insurance products from the reference insurance data. The list of candidate insurance products is a list of insurance that the target object has not purchased and serves as a first list that the target object may be interested in. Collaborative filtering and screening are performed based on the original insurance product characteristics to obtain a supplementary insurance product list as a second list that the target object may be interested in. Insurance products are recommended to the target object based on the candidate insurance product list and the supplementary insurance product list, thereby improving the efficiency and accuracy of insurance product recommendations.

[0144] The embodiment of the present application can make full use of the multi-dimensional information of customers and tap into potential needs through comprehensive data preprocessing and in-depth customer portrait construction, thereby achieving more accurate recommendations. The embodiment of the present application also establishes a real-time update and adjustment mechanism, which can optimize recommendation results based on market changes, product updates, and changes in customer needs to ensure the timeliness and effectiveness of recommendations. In addition, the embodiment of the present application combines two recommendation algorithms (user-based collaborative filtering and project-based coordinated filtering) and reasonably allocates weights to give full play to the advantages of various methods and improve the comprehensiveness and accuracy of recommendations.

[0145] Compared to the current intelligent recommendation method that relies on the user's basic information and limited purchase records, the embodiment of the present application can fully tap the user's potential needs and behavior patterns, improve the accuracy of insurance product recommendations, and fully value the user's feedback on the recommended insurance products, so as to timely optimize the recommendation strategy according to the user's actual response, and adapt to market changes and changes in customer needs through real-time updates and dynamic adjustments. In addition, the embodiment of the present application also pays attention to the protection gaps in the insurance products that the customer has purchased, and analyzes the protection gaps, so as to recommend insurance products that can fill the gaps in a targeted manner, providing customers with more complete risk protection. And through dimensionality reduction technology for visualization and interpretability: combining user tags and insurance product features for visual display, it increases the interpretability of the recommendation results, making it easier for customers to understand and accept the recommendations.

[0146] See also Figure 7 The present application also provides an insurance product recommendation device that can implement the above-mentioned insurance product recommendation method. The device includes:

[0147] Insurance data acquisition module, used to obtain the original insurance data of the target object;

[0148] A user profile construction module is used to construct a profile of the target object based on the original insurance data to obtain the original profile data of the target object; wherein the original profile data includes the original user label and the original insurance product features, and the original user label is multi-label;

[0149] The first collaborative filtering module is used to perform collaborative filtering based on the original user tags and the original insurance product features to obtain candidate objects;

[0150] An insurance product screening module is used to obtain reference insurance product data of a candidate subject and screen out a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is the data of insurance products purchased by the candidate subject, and the candidate insurance product list is a list of insurance products that the target subject has not purchased;

[0151] The second collaborative filtering module is used to perform collaborative filtering based on the characteristics of the original insurance products to obtain a list of supplementary insurance products;

[0152] The insurance product recommendation module is used to recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

[0153] In some embodiments, the user profile building module can be used to implement:

[0154] Extract key information from original insurance data to obtain original key information;

[0155] Perform feature encoding on the original key information to obtain preliminary encoding features;

[0156] The target object is clustered based on the preliminary coding features to obtain the cluster to which the target object belongs, the original user label and the original insurance product features.

[0157] Specifically, the tag acquisition module can be used to implement the above steps 201 to 203, which will not be described in detail here.

[0158] In some embodiments, the insurance product recommendation device may also be used to implement:

[0159] Configure the display colors for clusters according to the preset mapping configuration table;

[0160] Configure display identifiers for original user tags and original insurance product features according to the mapping configuration table;

[0161] The original portrait data is visualized according to the display color and display logo.

[0162] Specifically, the insurance product recommendation device can be used to implement the above steps 301 to 303, which will not be described in detail here.

[0163] In some embodiments, the second collaborative filtering module may be used to implement:

[0164] Obtaining preset insurance from a preset insurance resource pool, and performing feature extraction on the preset insurance to obtain preset insurance features;

[0165] Calculating similarity between the original insurance product features and the preset insurance features to obtain insurance product similarity between the original insurance product features and the preset insurance features;

[0166] A list of supplementary insurance products is selected from the insurance resource screening pool based on insurance product similarity.

[0167] Specifically, the second collaborative filtering module can be used to implement the above steps 401 to 403, which will not be described in detail here.

[0168] In some embodiments, the insurance product recommendation module may be used to implement:

[0169] screening a first selected insurance product from the list of candidate insurance products; wherein the first selected insurance product has a first weight;

[0170] screening a second selected insurance product from the list of supplementary insurance products; wherein the second selected insurance product has a second weight;

[0171] Performing weighted fusion on the first selected insurance product and the second selected insurance product based on the first weight and the second weight to obtain a list of selected insurance products;

[0172] Filter out target insurance products from the list of selected insurance products;

[0173] Recommend insurance products to target customers based on target insurance products.

[0174] Specifically, the insurance product recommendation module can be used to implement the above steps 501 to 505, which will not be described in detail here.

[0175] In some embodiments, the insurance product recommendation device may also be used to implement:

[0176] Obtain product information of a target insurance product and obtain target product information;

[0177] Generate insurance recommendation reports based on target insurance information and original insurance product features;

[0178] Push insurance recommendation reports to target objects.

[0179] Specifically, the insurance product recommendation device can be used to implement the above steps 601 to 603, which will not be described in detail here.

[0180] The specific implementation of the insurance product recommendation device is basically the same as the specific embodiment of the above-mentioned insurance product recommendation method, and will not be repeated here.

[0181] The present application also provides an electronic device comprising a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned insurance product recommendation method. The electronic device can be any smart terminal, such as a tablet computer or an in-vehicle computer.

[0182] See also Figure 8 , Figure 8 The hardware structure of an electronic device according to another embodiment is shown. The electronic device includes:

[0183] The processor 801 may be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application.

[0184] The memory 802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802 and is called by the processor 801 to execute the insurance product recommendation method of the embodiments of this application.

[0185] Input / output interface 803, used to implement information input and output;

[0186] Communication interface 804, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0187] Bus 805 , which transmits information between various components of the device (e.g., processor 801 , memory 802 , input / output interface 803 , and communication interface 804 );

[0188] The processor 801 , the memory 802 , the input / output interface 803 and the communication interface 804 are connected to each other in communication within the device via a bus 805 .

[0189] An embodiment of the present application further provides a storage medium, which is a computer-readable storage medium and stores a computer program. When the computer program is executed by a processor, the above-mentioned insurance product recommendation method is implemented.

[0190] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0191] The insurance product recommendation method and device, electronic device, and storage medium provided in the embodiments of the present application can construct a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, which includes the original user label and the original insurance product characteristics. Collaborative filtering and screening are performed based on the original user label and the original insurance product characteristics to obtain candidate objects similar to the target object, obtain reference insurance product data of the candidate objects, and screen out a list of candidate insurance products from the reference insurance data. The list of candidate insurance products is a list of insurance that the target object has not purchased and serves as a first list that the target object may be interested in. Collaborative filtering and screening are performed based on the original insurance product characteristics to obtain a supplementary insurance product list as a second list that the target object may be interested in. Insurance products are recommended to the target object based on the candidate insurance product list and the supplementary insurance product list, thereby improving the efficiency and accuracy of insurance product recommendations and better matching user needs. Compared to the current intelligent recommendation method that relies on the user's basic information and limited purchase records, the embodiment of the present application can fully tap the user's potential needs and behavior patterns, improve the accuracy of insurance product recommendations, and fully value the user's feedback on the recommended insurance products, so as to timely optimize the recommendation strategy according to the user's actual response, and adapt to market changes and changes in customer needs through real-time updates and dynamic adjustments. In addition, the embodiment of the present application also pays attention to the protection gaps in the insurance products that the customer has purchased, and analyzes the protection gaps, so as to recommend insurance products that can fill the gaps in a targeted manner, providing customers with more complete risk protection. And through dimensionality reduction technology for visualization and interpretability: combining user tags and insurance product features for visual display, it increases the interpretability of the recommendation results, making it easier for customers to understand and accept the recommendations.

[0192] The embodiments described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0193] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0194] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0195] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0196] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0197] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0198] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the above-mentioned units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0199] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0200] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0201] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes multiple instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: various media that can store programs, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0202] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but are not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present invention should be within the scope of the present invention.

Claims

1. A method for recommending insurance products, characterized in that: The method comprises: Obtain the original insurance data of the target object; Constructing a profile of the target object based on the original insurance data to obtain original profile data of the target object; wherein the original profile data includes original user tags and original insurance product features, and the original user tags are multiple tags; Perform collaborative filtering based on the original user tags and the original insurance product features to obtain candidate objects; Obtain reference insurance product data of the candidate subject, and filter out a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is data of insurance products purchased by the candidate subject, and the list of candidate insurance products is a list of insurance products not purchased by the target subject; Perform collaborative filtering based on the characteristics of the original insurance products to obtain a list of supplementary insurance products; Recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

2. The method according to claim 1, characterized in that The collaborative filtering based on the original insurance product features is performed to obtain a list of supplementary insurance products, including: Obtaining preset insurance from a preset insurance resource pool, and performing feature extraction on the preset insurance to obtain preset insurance features; Calculating similarity between the original insurance product features and the preset insurance features to obtain insurance product similarity between the original insurance product features and the preset insurance features; The supplementary insurance product list is selected from the insurance resource screening pool based on the insurance product similarity.

3. The method according to claim 1, characterized in that The recommending insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list includes: Screening a first selected insurance product from the candidate insurance product list; wherein the first selected insurance product has a first weight; Filtering a second selected insurance product from the list of supplementary insurance products; wherein the second selected insurance product has a second weight; Performing weighted calculation on the first selected insurance product and the second selected insurance product based on the first weight and the second weight to obtain a list of selected insurance products; Filtering target insurance products from the selected insurance product list; Recommend insurance products to the target object based on the target insurance product.

4. The method according to claim 3, characterized in that After recommending an insurance product to the target object based on the target insurance product, the method further includes: pushing a recommendation report, specifically including: Obtaining product information of the target insurance product to obtain target product information; generating an insurance recommendation report based on the target insurance information and the characteristics of the original insurance product; Push the insurance recommendation report to the target object.

5. The method according to any one of claims 2 to 4, characterized in that The original portrait data also includes clusters. The portrait of the target object is constructed based on the original insurance data to obtain the original portrait data of the target object, including: Extracting key information from the original insurance data to obtain original key information; Performing feature encoding on the original key information to obtain preliminary encoding features; The target object is clustered based on the preliminary coding features to obtain the cluster to which the target object belongs, the original user label, and the original insurance product features.

6. The method according to claim 5, characterized in that After constructing a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, the method further includes: Configuring display colors for the clusters according to a preset mapping configuration table; Configuring a display identifier for the original user tag and the original insurance product feature according to the mapping configuration table; The original portrait data is visually displayed according to the display color and the display identifier.

7. The method according to claim 5, characterized in that After constructing a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object, the method further includes: updating data, specifically including: In response to a periodic update request, collecting new insurance data at a preset period; wherein the new insurance data includes new insurance products; updating the insurance resource pool according to the new insurance data; wherein the new insurance product is used as the preset insurance; The original portrait data is updated based on the updated insurance resource pool.

8. An insurance product recommendation device, characterized in that: The device comprises: Insurance data acquisition module, used to obtain the original insurance data of the target object; A user portrait construction module is used to construct a portrait of the target object based on the original insurance data to obtain the original portrait data of the target object; wherein the original portrait data includes the original user label and the original insurance product feature, and the original user label is a multi-label; A first collaborative filtering module is configured to perform collaborative filtering based on the original user tags and the original insurance product features to obtain candidate objects; An insurance product screening module, configured to obtain reference insurance product data of the candidate subject and screen a list of candidate insurance products from the reference insurance data; wherein the reference insurance product data is data of insurance products purchased by the candidate subject, and the list of candidate insurance products is a list of insurance products not purchased by the target subject; A second collaborative filtering module is configured to perform collaborative filtering based on the characteristics of the original insurance products to obtain a list of supplementary insurance products; An insurance product recommendation module is used to recommend insurance products to the target object based on the candidate insurance product list and the supplementary insurance product list.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images