Product recommendation method and device, computer equipment and storage medium
By extracting user feature vectors from social media information, and filtering and sorting promotional information for insurance products, the problem of low customer acquisition efficiency for insurance companies has been solved, achieving precise marketing and efficient customer acquisition.
Patent Information
- Application Number
- CN202511071831.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-31
- Publication Date
- 2025-11-18
AI Technical Summary
Insurance companies lack in-depth research into customers' potential needs and health conditions during the customer acquisition process, resulting in high customer acquisition costs, low conversion rates, and inaccurate customer profiles.
By acquiring users' social media information, determining user feature vectors, filtering target insurance products, and generating the display order of product promotional information based on user feature vectors, precise marketing can be achieved.
It improved the efficiency of customer acquisition for insurance products and the accuracy of customer profiles, reduced customer acquisition costs, and increased the conversion rate.
Smart Images
Figure CN120975929A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of financial services, and more particularly to a product recommendation method, apparatus, computer equipment, and storage medium. Background Technology
[0002] Big data can help businesses gain a deeper understanding of consumer behavior patterns, preferences, and needs. By analyzing consumer purchase history, browsing behavior, and social media interactions, businesses can achieve precise advertising and personalized recommendations. For example, e-commerce platforms use big data to analyze consumer shopping habits and recommend products that users may be interested in, thereby improving sales conversion rates and customer satisfaction. However, insurance products are diverse, including medical insurance, property insurance, and financial insurance, and user preference data for insurance products is characterized by scattered sources and difficulty in integration and analysis. Currently, insurance companies still rely mainly on traditional methods such as telemarketing, offline visits, and social media advertising for customer acquisition, lacking in-depth analysis of customers' potential needs and health conditions, resulting in high customer acquisition costs, low conversion rates, and inaccurate customer profiles. Summary of the Invention
[0003] The main purpose of this application is to provide a product recommendation method, apparatus, device, and computer storage medium, aimed at improving the customer acquisition efficiency of insurance products.
[0004] Firstly, this application provides a product recommendation method, which includes the following steps:
[0005] Obtain the social media information corresponding to the user, and determine the user's feature vector based at least on the social media information;
[0006] Based on the user feature vector, multiple target insurance products are selected from the preset candidate insurance products;
[0007] Generate product promotion information for the target insurance product based on the user feature vector;
[0008] The display order of the product promotional information is determined based on the target matching degree between the user feature vector and the target insurance product, and the product promotional information is displayed according to the display order.
[0009] Secondly, this application also provides a product recommendation device, the product recommendation device comprising:
[0010] The user feature extraction module is used to obtain social media information corresponding to a user, and to determine the user feature vector based at least on the social media information.
[0011] The target product screening module is used to screen multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector.
[0012] The promotional information generation module is used to generate promotional information for the target insurance product based on the user feature vector.
[0013] The promotional information display module is used to determine the display order of the product promotional information based on the matching degree between the user feature vector and the target insurance product, and to display the product promotional information according to the display order.
[0014] Thirdly, this application also provides a computer device, the computer device including a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the product recommendation method as described above.
[0015] Fourthly, this application also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the product recommendation method as described above.
[0016] This application provides a product recommendation method, apparatus, device, and computer storage medium. The method involves acquiring a user's social media information and determining the user's user feature vector based on the social media information. Based on the user feature vector, multiple target insurance products are selected from a preset pool of candidate insurance products. Promotional information for the target insurance products is generated based on the user feature vector. The display order of the promotional information is determined based on the target matching degree between the user feature vector and the target insurance products, and the promotional information is displayed according to the specified order. By identifying target insurance products and generating promotional information based on social media information, and providing customers with medical insurance, property insurance, and financial insurance products according to actual needs, precise marketing of insurance products is achieved, improving customer acquisition efficiency. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 A schematic flowchart illustrating a product recommendation method provided in an embodiment of this application;
[0019] Figure 2This application provides a usage scenario diagram of a product recommendation method according to an embodiment of the present application;
[0020] Figure 3 A schematic block diagram of a product recommendation device provided in an embodiment of this application;
[0021] Figure 4 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application. Detailed Implementation
[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0023] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.
[0024] This application provides a product recommendation method, apparatus, computer device, and computer-readable storage medium.
[0025] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.
[0026] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a product recommendation method provided in an embodiment of this application. This product recommendation method can be used in a terminal or server to identify a target insurance product and generate and display promotional information for that target insurance product. The terminal can be an electronic device such as a mobile phone, tablet, laptop, desktop computer, personal digital assistant, or wearable device; the server can be a standalone server, a server cluster, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.
[0027] Please refer to Figure 2 , Figure 2 This is a usage scenario diagram provided by an embodiment of this application. For example... Figure 2As shown, user feature vectors are obtained by extracting features from users' social media information. Target insurance products are then selected from candidate insurance products based on these user feature vectors, and promotional information for the target insurance products is generated based on the user feature vectors.
[0028] like Figure 1 As shown, the recommended method for this product includes steps S101 to S104.
[0029] Step S101: Obtain the social media information corresponding to the user, and determine the user feature vector based at least on the social media information.
[0030] For example, the content posted and interacted by users on social media platforms can reflect their consumption preferences, which can then serve as a basis for recommending insurance products to users. Therefore, feature extraction can be performed on social media information to obtain user feature vectors.
[0031] Of course, this is not the only possibility. User feature vectors can also be determined by combining social media information with at least one of the following: intelligent medical consultation information and health monitoring information. Among them, intelligent medical consultation information can be obtained from the user's chronic disease history, family medical history, health problems, etc., through a natural language processing model when the user is on the insurance product selection interface; health monitoring information can be data monitored by health monitoring devices such as watches and bracelets worn by the user.
[0032] In some embodiments, after displaying the product promotional information according to the display order, the method further includes:
[0033] After receiving a notification message from a user about purchasing an insurance product, the system obtains the user's health information, which includes at least: intelligent consultation information and health monitoring information.
[0034] Based on the health information, determine the user's health risk type, and based on the health risk type, determine the health advice information to be displayed to the user;
[0035] Based on the health risk type and the user feature vector, multiple target insurance products are selected from the preset candidate insurance products.
[0036] For example, after a user purchases an insurance product based on product promotional information, intelligent consultation information can be obtained through a large language model, such as the user's potential health problems, chronic disease history, family medical history, etc. Based on the intelligent consultation information, it can be identified whether the user is in a high-risk state such as hypertension, diabetes, cardiovascular and cerebrovascular diseases; and by establishing a communication connection with the user's smart wearable device, health monitoring information can be obtained, such as the user's heart rate information, sleep quality information, and exercise amount information.
[0037] For example, users can be clustered based on the aforementioned health information. For instance, multiple population groups can be pre-defined as preset risk types, or multiple risk levels can be pre-defined as preset risk types. The user's health information can be compared with each preset risk type to determine the preset risk type to which the user belongs as the health risk type.
[0038] For example, based on the user's type of health risk, health advice can be provided, such as reminding the user to adjust their diet and lifestyle, exercise regularly, and have regular check-ups, thereby intervening in the user's health, combining disease prevention with protection, improving the user's health level, and thus reducing the claim rate.
[0039] Specifically, after a user purchases insurance, their health information can be continuously monitored through smart wearable devices such as fitness trackers and smartwatches. If any abnormal indicators appear in the user's health monitoring information (such as abnormal heart rate, decreased sleep quality, or a sudden decrease in exercise), potential claims risks can be warned in advance so that the user's policy can be reinsured; and relevant health advice information can be pushed to the user.
[0040] For example, after obtaining the health risk type, new target insurance products suitable for the user can be determined by combining the health risk type and user feature vector, thereby improving customer retention and user stickiness. For instance, critical illness insurance and medical insurance can be recommended to users with hereditary disease risks or those in a long-term sub-healthy state, while long-term care insurance can be provided to users with care needs. Specifically, the correspondence between each health risk type and candidate insurance products can be pre-defined, thereby determining the target insurance product from the candidate insurance products based on the user's health risk type. Of course, this is not limited to this approach and is not specified here.
[0041] In some implementations, obtaining the user's corresponding social media information and determining the user's feature vector based at least on the social media information includes:
[0042] In response to user-authorized user information, at least one piece of publicly available social media information from the platform is obtained based on the user information, the social media information including posted information and interactive information;
[0043] Feature extraction is performed on the published information and the interactive information to obtain the user feature vector.
[0044] For example, to avoid infringing on user privacy, obtaining social media information requires user authorization. For instance, users can authorize the insurance product filtering interface by entering their IDs on various social media platforms. Furthermore, the obtained social media information is limited to publicly available information on the social media platforms, such as user activity on those platforms that can be accessed through guest mode.
[0045] For example, a user's social media information on a social platform includes posted information and interactive information. User information is content actively posted by the user, such as asking questions, answering questions, posting updates, and commenting. Interactive information is content in which the user participates in interactions, such as following or liking.
[0046] In some implementations, the step of extracting features from the published information and the interactive information to obtain the user feature vector includes:
[0047] Feature extraction is performed on the published information to obtain a published information sub-vector, and feature extraction is performed on the interactive information to obtain an interactive information sub-vector;
[0048] The user feature vector is obtained by fusing features from the published information subvector and the interactive information subvector.
[0049] Wherein, the weight of the first vector corresponding to the published information sub-vector is greater than the weight of the second vector corresponding to the interactive information sub-vector.
[0050] Understandably, in social media messaging, posted information reflects user preferences more effectively than interactive information; therefore, posted information is given higher weight.
[0051] Specifically, features can be extracted from published information and interactive information separately, and then the extracted published information sub-vectors and interactive information sub-vectors can be fused based on different weights to obtain a user feature vector, and the weight of the first vector corresponding to the published information sub-vector is greater than the weight of the second vector corresponding to the interactive information sub-vector.
[0052] Step S102: Based on the user feature vector, select multiple target insurance products from the preset candidate insurance products.
[0053] For example, since user feature vectors reflect users' consumption preferences, they can be compared with the product feature vectors corresponding to candidate insurance products to obtain the similarity between the user feature vectors and product feature vectors (e.g., cosine similarity). The target insurance product can then be selected from the candidate insurance products based on the magnitude of the similarity. The product feature vector can be determined based on the user feature vectors of users who purchase the insurance product, but it is not limited to this; the product feature vector can also be obtained by extracting features from product-related promotional information.
[0054] In some implementations, the user feature vector includes a user preference vector and a user consumption vector; the step of selecting multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector includes:
[0055] The preference matching degree is determined based on the user preference vector and the product preference vector of the candidate insurance products.
[0056] The price matching degree is determined based on the user consumption vector and the product price vector of the candidate insurance products;
[0057] The target matching degree of the candidate insurance products is determined based on the preference matching degree and the price matching degree, and the target insurance product is selected from the candidate insurance products based on the target matching degree.
[0058] For example, the user preference vector is used to reflect user preferences. For instance, users who enjoy sports may tend to buy accident insurance, while users who are concerned about their children's education may tend to buy education savings insurance. The user consumption vector is used to reflect the user's consumption level, so that insurance products at different price points can be recommended to the user based on their consumption level.
[0059] For example, the product feature vector may also include a product preference vector and a product price vector, wherein the product preference vector is used to reflect the characteristics of the target population for which the insurance product is targeted, and the product price vector is used to reflect the price level of the insurance product.
[0060] For example, the preference matching degree is obtained by comparing the user's preference vector with the product preference vector of each candidate insurance product, and the price matching degree is obtained by comparing the user's consumption vector with the product price vector of each candidate insurance product. The target matching degree between the user and the candidate insurance products is then determined by combining the preference matching degree and the price matching degree. The preference matching degree and price matching degree can be, for example, cosine similarity, but are not limited to this, and are not restricted here.
[0061] In some implementations, determining the target matching degree of the candidate insurance product based on the preference matching degree and the price matching degree includes:
[0062] The user's price sensitivity is determined based on the social media information;
[0063] The first matching weight and the second matching weight of the price matching degree are determined based on the price sensitivity.
[0064] A first matching degree is determined based on the preference matching degree and the corresponding first matching degree weight, and a second matching degree is determined based on the price matching degree and the corresponding second matching degree weight. The sum of the first matching degree and the second matching degree is taken as the target matching degree.
[0065] For example, different users have different price sensitivities. Some users may focus more on the price of insurance products, while others may pay more attention to other aspects of the insurance product. Therefore, a user's price sensitivity can be determined based on their social media information, such as by determining the user's price sensitivity based on their consumption vector. Specifically, the user's consumption vector is input into a sensitivity partitioning model to obtain price sensitivity. The sensitivity partitioning model can be implemented based on models such as decision trees and random forests, which will not be elaborated on here.
[0066] Of course, it is not limited to this. Users can also be clustered with multiple preset groups based on social media information to obtain the category mentioned by the user. The price sensitivity of the group can then be used as the user's price sensitivity. This is not limited here.
[0067] In some implementations, determining the first matching weight and the second matching weight of the price matching degree based on the price sensitivity includes:
[0068] The second matching degree weight is determined based on the price sensitivity according to the preset correspondence, and the second matching degree weight is positively correlated with the price sensitivity;
[0069] The first matching degree weight is determined based on the second matching degree weight.
[0070] For example, for users who pay more attention to price, a higher weight can be assigned to price matching to increase the proportion of price matching in the target matching; conversely, for users who pay less attention to price, a lower weight can be assigned to price matching to increase the proportion of preference matching in the target matching.
[0071] For example, the first matching degree weight can be obtained by subtracting the second matching degree weight from 1, while the target matching degree can be obtained by adding the product of the price matching degree and the second matching degree weight and the product of the preference matching degree and the first matching degree weight.
[0072] Step S103: Generate product promotion information for the target insurance product based on the user feature vector.
[0073] For example, the product recommendation method provided in this application generates personalized promotional information for target insurance products for users based on user feature vectors, thereby matching the promotional information with the user's needs, clearly and intuitively displaying the highlights of the target insurance products to the user, and improving the customer acquisition success rate and efficiency of the target insurance products.
[0074] In some implementations, generating product promotional information for the target insurance product based on the user feature vector includes:
[0075] If the first matching degree is greater than the second matching degree, the product promotion information is generated based on the product preference vector of the target insurance product;
[0076] If the first matching degree is less than or equal to the second matching degree, the product promotion information is generated based on the product price vector of the target insurance product.
[0077] For example, product promotional information may be image information, text information, etc., obtained by inputting the product feature vector of the target insurance product into a preset natural language big data model.
[0078] For example, for a target insurance product with a first matching degree greater than a second matching degree, it means that the other features of the insurance product are more attractive to users than the price. Therefore, the product preference vector is used as the basis for generating product promotional information. Conversely, for a target insurance product with a first matching degree less than or equal to a second matching degree, it means that the price of the insurance product is more attractive to users than the other features. Therefore, the product price vector is used as the basis for generating product promotional information.
[0079] Step S104: Determine the display order of the product promotional information based on the target matching degree between the user feature vector and the target insurance product, and display the product promotional information according to the display order.
[0080] For example, the display order of product promotional information can be determined based on the degree of target matching, so that the promotional information of target insurance products with higher target matching is displayed first, increasing the likelihood that users will pay attention to products they are interested in, thereby increasing the customer acquisition success rate.
[0081] The product recommendation method provided in the above embodiments obtains the user's corresponding social media information and determines the user's user feature vector based at least on the social media information; selects multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector; generates promotional information for the target insurance products based on the user feature vector; determines the display order of the promotional information based on the target matching degree between the user feature vector and the target insurance products; and displays the promotional information according to the display order. By determining target insurance products for users and generating promotional information based on social media information, and providing customers with insurance products such as medical insurance, property insurance, and financial insurance according to actual needs, precise marketing of insurance products is achieved, improving customer acquisition efficiency for insurance products.
[0082] Please see Figure 3 , Figure 3This is a schematic diagram of a product recommendation device provided in an embodiment of this application. The product recommendation device can be configured in a server or terminal to execute the aforementioned product recommendation method.
[0083] like Figure 3 As shown, the product recommendation device includes: a user feature extraction module 110, a target product filtering module 120, a promotional information generation module 130, and a promotional information display module 140.
[0084] User feature extraction module 110 is used to obtain social media information corresponding to a user and determine the user feature vector of the user based at least on the social media information;
[0085] The target product screening module 120 is used to screen multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector.
[0086] The promotional information generation module 130 is used to generate product promotional information for the target insurance product based on the user feature vector.
[0087] The promotional information display module 140 is used to determine the display order of the product promotional information based on the matching degree between the user feature vector and the target insurance product, and to display the product promotional information according to the display order.
[0088] In some embodiments, the product recommendation device is also used to achieve:
[0089] Obtaining users' health information, which includes at least: intelligent consultation information and health monitoring information;
[0090] Based on the health information, determine the user's health risk type, and based on the health risk type, determine the health advice information to be displayed to the user;
[0091] In the process of selecting multiple target insurance products from preset candidate insurance products based on the user feature vector, the target product screening module 120 is used to:
[0092] Based on the health risk type and the user feature vector, multiple target insurance products are selected from the preset candidate insurance products.
[0093] In some implementations, the user feature extraction module 110, in the process of acquiring the social media information corresponding to the user and determining the user feature vector based at least on the social media information, is used to:
[0094] In response to user-authorized user information, at least one piece of publicly available social media information from the platform is obtained based on the user information, the social media information including posted information and interactive information;
[0095] Feature extraction is performed on the published information and the interactive information to obtain the user feature vector.
[0096] In some implementations, the user feature extraction module 110, in the process of extracting features from the published information and the interaction information to obtain the user feature vector, is used to:
[0097] Feature extraction is performed on the published information to obtain a published information sub-vector, and feature extraction is performed on the interactive information to obtain an interactive information sub-vector;
[0098] The user feature vector is obtained by fusing features from the published information subvector and the interactive information subvector.
[0099] Wherein, the weight of the first vector corresponding to the published information sub-vector is greater than the weight of the second vector corresponding to the interactive information sub-vector.
[0100] In some implementations, the target product screening module 120, in the process of screening multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector, is used to:
[0101] The preference matching degree is determined based on the user preference vector and the product preference vector of the candidate insurance products.
[0102] The price matching degree is determined based on the user consumption vector and the product price vector of the candidate insurance products;
[0103] The target matching degree of the candidate insurance products is determined based on the preference matching degree and the price matching degree, and the target insurance product is selected from the candidate insurance products based on the target matching degree.
[0104] In some implementations, the target product screening module 120, in the process of determining the target matching degree of the candidate insurance products based on the preference matching degree and the price matching degree, is used to:
[0105] The user's price sensitivity is determined based on the social media information;
[0106] A first matching weight and a second matching weight of the price matching degree are determined based on the price sensitivity, wherein the second matching weight is positively correlated with the price sensitivity;
[0107] A first matching degree is determined based on the preference matching degree and the corresponding first matching degree weight, and a second matching degree is determined based on the price matching degree and the corresponding second matching degree weight. The sum of the first matching degree and the second matching degree is taken as the target matching degree.
[0108] In some implementations, the promotional information generation module 130, in the process of generating product promotional information for the target insurance product based on the user feature vector, is used to:
[0109] If the first matching degree is greater than the second matching degree, the product promotion information is generated based on the product preference vector of the target insurance product;
[0110] If the first matching degree is less than or equal to the second matching degree, the product promotion information is generated based on the product price vector of the target insurance product.
[0111] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the above-described apparatus and its modules and units can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0112] The methods and apparatus of this application can be used in a wide variety of general-purpose or special-purpose computing system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This 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, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0113] For example, the above-described method and apparatus can be implemented as a computer program, which can be used in, for example... Figure 4 It runs on the computer device shown.
[0114] Please see Figure 4 , Figure 4 This is a schematic block diagram illustrating the structure of a computer device provided in an embodiment of this application. The computer device may be a server or a terminal.
[0115] like Figure 4 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a storage medium and internal memory.
[0116] The storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any recommended product method.
[0117] The processor provides computing and control capabilities, supporting the operation of the entire computer device.
[0118] Internal memory provides an environment for the execution of computer programs stored in storage media. When these computer programs are executed by a processor, the processor can perform any product recommendation method.
[0119] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0120] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.
[0121] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps:
[0122] Obtain the social media information corresponding to the user, and determine the user's feature vector based at least on the social media information;
[0123] Based on the user feature vector, multiple target insurance products are selected from the preset candidate insurance products;
[0124] Generate product promotion information for the target insurance product based on the user feature vector;
[0125] The display order of the product promotional information is determined based on the target matching degree between the user feature vector and the target insurance product, and the product promotional information is displayed according to the display order.
[0126] In some embodiments, the processor, in implementing the product recommendation method, is further configured to:
[0127] Obtaining users' health information, which includes at least: intelligent consultation information and health monitoring information;
[0128] Based on the health information, determine the user's health risk type, and based on the health risk type, determine the health advice information to be displayed to the user;
[0129] In some implementations, the processor, in the process of selecting multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector, is configured to:
[0130] Based on the health risk type and the user feature vector, multiple target insurance products are selected from the preset candidate insurance products.
[0131] In some implementations, the processor, in the process of acquiring the user's corresponding social media information and determining the user's feature vector based at least on the social media information, is configured to:
[0132] In response to user-authorized user information, at least one piece of publicly available social media information from the platform is obtained based on the user information, the social media information including posted information and interactive information;
[0133] Feature extraction is performed on the published information and the interactive information to obtain the user feature vector.
[0134] In some implementations, the processor, in the process of extracting features from the published information and the interaction information to obtain the user feature vector, is used to:
[0135] Feature extraction is performed on the published information to obtain a published information sub-vector, and feature extraction is performed on the interactive information to obtain an interactive information sub-vector;
[0136] The user feature vector is obtained by fusing features from the published information subvector and the interactive information subvector.
[0137] Wherein, the weight of the first vector corresponding to the published information sub-vector is greater than the weight of the second vector corresponding to the interactive information sub-vector.
[0138] In some implementations, the processor, in implementing the user feature vector including a user preference vector and a user consumption vector; and in the process of selecting multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector, is used to:
[0139] The preference matching degree is determined based on the user preference vector and the product preference vector of the candidate insurance products.
[0140] The price matching degree is determined based on the user consumption vector and the product price vector of the candidate insurance products;
[0141] The target matching degree of the candidate insurance products is determined based on the preference matching degree and the price matching degree, and the target insurance product is selected from the candidate insurance products based on the target matching degree.
[0142] In some implementations, the processor, in the process of determining the target matching degree of the candidate insurance product based on the preference matching degree and the price matching degree, is configured to:
[0143] The user's price sensitivity is determined based on the social media information;
[0144] A first matching weight and a second matching weight of the price matching degree are determined based on the price sensitivity, wherein the second matching weight is positively correlated with the price sensitivity;
[0145] A first matching degree is determined based on the preference matching degree and the corresponding first matching degree weight, and a second matching degree is determined based on the price matching degree and the corresponding second matching degree weight. The sum of the first matching degree and the second matching degree is taken as the target matching degree.
[0146] In some implementations, the processor, in the process of generating product promotional information for the target insurance product based on the user feature vector, is configured to:
[0147] If the first matching degree is greater than the second matching degree, the product promotion information is generated based on the product preference vector of the target insurance product;
[0148] If the first matching degree is less than or equal to the second matching degree, the product promotion information is generated based on the product price vector of the target insurance product.
[0149] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the computer device described above can be referred to the corresponding process in the aforementioned product recommendation method embodiments, and will not be repeated here.
[0150] This application also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to various embodiments of the product recommendation method of this application.
[0151] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.
[0152] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.
[0153] It should also be understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. It should be noted that, herein, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.
[0154] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. The above descriptions are merely specific implementations of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A product recommendation method, characterized in that, The method includes: Obtain the social media information corresponding to the user, and determine the user's feature vector based at least on the social media information; Based on the user feature vector, multiple target insurance products are selected from the preset candidate insurance products; Generate product promotion information for the target insurance product based on the user feature vector; The display order of the product promotional information is determined based on the target matching degree between the user feature vector and the target insurance product, and the product promotional information is displayed according to the display order.
2. The product recommendation method according to claim 1, characterized in that, After displaying the product promotional information according to the display order, the method further includes: After receiving a notification message from a user about purchasing an insurance product, the system obtains the user's health information, which includes at least: intelligent consultation information and health monitoring information. Based on the health information, determine the user's health risk type, and based on the health risk type, determine the health advice information to be displayed to the user; Based on the health risk type and the user feature vector, multiple target insurance products are selected from the preset candidate insurance products.
3. The product recommendation method according to claim 1, characterized in that, The step of obtaining the user's corresponding social media information and determining the user's user feature vector based at least on the social media information includes: In response to user-authorized user information, at least one piece of publicly available social media information from the platform is obtained based on the user information, the social media information including posted information and interactive information; Feature extraction is performed on the published information and the interactive information to obtain the user feature vector.
4. The product recommendation method according to claim 3, characterized in that, The step of extracting features from the published information and the interactive information to obtain the user feature vector includes: Feature extraction is performed on the published information to obtain a published information sub-vector, and feature extraction is performed on the interactive information to obtain an interactive information sub-vector; The user feature vector is obtained by fusing features from the published information subvector and the interactive information subvector. Wherein, the weight of the first vector corresponding to the published information sub-vector is greater than the weight of the second vector corresponding to the interactive information sub-vector.
5. The product recommendation method according to claim 1, characterized in that, The user feature vector includes a user preference vector and a user consumption vector; the step of selecting multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector includes: The preference matching degree is determined based on the user preference vector and the product preference vector of the candidate insurance products. The price matching degree is determined based on the user consumption vector and the product price vector of the candidate insurance products; The target matching degree of the candidate insurance products is determined based on the preference matching degree and the price matching degree, and the target insurance product is selected from the candidate insurance products based on the target matching degree.
6. The product recommendation method according to claim 5, characterized in that, Determining the target matching degree of the candidate insurance products based on the preference matching degree and the price matching degree includes: The user's price sensitivity is determined based on the social media information; A first matching weight and a second matching weight of the price matching degree are determined based on the price sensitivity, wherein the second matching weight is positively correlated with the price sensitivity; A first matching degree is determined based on the preference matching degree and the corresponding first matching degree weight, and a second matching degree is determined based on the price matching degree and the corresponding second matching degree weight. The sum of the first matching degree and the second matching degree is taken as the target matching degree.
7. The product recommendation method according to claim 6, characterized in that, The step of generating product promotional information for the target insurance product based on the user feature vector includes: If the first matching degree is greater than the second matching degree, the product promotion information is generated based on the product preference vector of the target insurance product; If the first matching degree is less than or equal to the second matching degree, the product promotion information is generated based on the product price vector of the target insurance product.
8. A product recommendation device, characterized in that, The product recommendation device includes: The user feature extraction module is used to obtain social media information corresponding to a user, and to determine the user feature vector based at least on the social media information. The target product screening module is used to screen multiple target insurance products from a preset pool of candidate insurance products based on the user feature vector. The promotional information generation module is used to generate promotional information for the target insurance product based on the user feature vector. The promotional information display module is used to determine the display order of the product promotional information based on the matching degree between the user feature vector and the target insurance product, and to display the product promotional information according to the display order.
9. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of the product recommendation method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the product recommendation method as described in any one of claims 1 to 7.