Service recommendation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a service recommendation method and device, electronic equipment and a storage medium, and is applied to the technical field of service processing, and the method comprises the steps: firstly obtaining sample services and features of a plurality of users, and determining the features of a target user; and selecting target sample services according to the number, and taking the service values of the target sample services as centroids. And distributing other sample services into the cluster according to the similarity. And determining the service mean value of each cluster, and if the difference between the centroid and the mean value reaches a preset value, executing the step of repeatedly distributing by taking the mean value as a new centroid until the centroid is stable. And finally, converting the centroid into a recommendation service, and providing a corresponding recommendation service for each target user feature. Therefore, the service clustering is performed on the sample service according to different user characteristics in advance, and different recommended services are determined for different user characteristics according to the clustering result, so that the corresponding recommended services can be determined subsequently according to the user characteristics, service recommendation is performed, and the service recommendation efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of business processing technology, and in particular to a business recommendation method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of the financial insurance field, especially in the current era of rapid development of the Internet economy, the traditional insurance underwriting system can no longer meet the expansion of actual insurance business. At present, the traditional insurance quotation method is to expand business through salesmen, telephone sales and customers purchasing insurance by themselves. All three aspects are inseparable from the quotation of insurance policies.

[0003] In the related art, without understanding the customer's purchasing preferences, the salesperson needs to communicate with the customer multiple times in order to modify the insurance policy and make a quotation. Each modification of the content of the insurance policy requires manual operation by the salesperson, which is time-consuming and labor-intensive. Therefore, it is urgent to provide a business recommendation method to solve the above technical problems. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to propose a business recommendation method, device, electronic device and storage medium, which can determine different recommended businesses according to different user characteristics in advance, so as to facilitate subsequent business recommendations based on user characteristics, avoid multiple modifications to the insurance policy, and improve business recommendation efficiency.

[0005] To achieve the above object, a first aspect of an embodiment of the present application proposes a service recommendation method, the method comprising:

[0006] Acquire sample services and user characteristics of multiple users, and determine target user characteristics from the multiple user characteristics;

[0007] According to the number of features of the target user features, a target sample service is selected from the plurality of sample services, and the service value of the target sample service is respectively used as the centroid of each cluster corresponding to the target user feature;

[0008] Based on the similarity between the service values ​​of other sample services and each centroid, allocating other sample services to the cluster where the corresponding centroid is located, wherein the other sample services are services other than the centroid in the sample services;

[0009] Determine the business mean of each cluster;

[0010] When there is a target cluster in multiple clusters whose difference between the centroid and the corresponding service mean reaches a preset difference value, the service mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between other sample services and each centroid is returned to be executed, until there is no target cluster in multiple clusters whose difference between the centroid and the corresponding service mean reaches a preset difference value, and the centroid of each cluster corresponding to the target user feature is obtained;

[0011] The centroid of the cluster corresponding to each target user feature is converted into a corresponding recommended service to obtain a recommended service corresponding to each target user feature.

[0012] In some embodiments, allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between the service value of other sample services and each centroid includes:

[0013] Obtain the similarity between the business values ​​of other sample businesses and each centroid;

[0014] Assign other sample businesses to the cluster where the target centroid with the smallest similarity to each centroid is located.

[0015] In some embodiments, obtaining the similarity between the service values ​​of other sample services and each centroid includes:

[0016] For each centroid, obtain the first sub-service value corresponding to each service type in other sample services, and obtain the second sub-service value corresponding to each service type in the centroid;

[0017] Calculate the difference between the first sub-service value and the second sub-service value corresponding to the same service type to obtain the sub-service difference corresponding to each service type;

[0018] Based on the sub-service difference corresponding to each service type, the similarity between the service values ​​of other sample services and each centroid is determined.

[0019] In some embodiments, determining the similarity between the service values ​​of other sample services and each centroid based on the sub-service difference corresponding to each service type includes:

[0020] Calculate the square of the sub-service difference corresponding to each service type to obtain a first calculation result;

[0021] Calculate the sum of multiple first calculation results to obtain a second calculation result;

[0022] The second calculation result is subjected to square root processing to obtain the similarity between the business values ​​of other sample businesses and each centroid.

[0023] In some embodiments, determining the service mean of each cluster includes:

[0024] For each cluster, obtain the business volume of the services in the cluster and the business value corresponding to each service;

[0025] Calculate the sum of the service values ​​corresponding to each service to obtain a third calculation result;

[0026] A ratio of the third calculation result to the traffic volume is calculated to obtain a traffic mean of each cluster.

[0027] In some embodiments, converting the centroid of the cluster corresponding to each target user feature into a corresponding recommended service to obtain the recommended service corresponding to each target user feature includes:

[0028] Determine the target service value interval in which the centroid of each cluster corresponding to the target user feature is located;

[0029] The target service corresponding to each target service value interval is determined as a recommended service corresponding to the target user characteristics.

[0030] In some embodiments, after converting the centroid of the cluster corresponding to each target user feature into the corresponding recommended service to obtain the recommended service corresponding to each target user feature, the method further includes:

[0031] Obtain the recommended user characteristics of the recommended user;

[0032] Filtering out a matching user feature that matches the user feature to be recommended from the plurality of target user features;

[0033] Recommend the recommended service corresponding to the matching user feature to the to-be-recommended user.

[0034] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application proposes a service recommendation device, including:

[0035] A first acquisition unit is used to acquire sample services and user features of multiple users, and determine target user features from the multiple user features;

[0036] A selection unit, configured to select a target sample service from the plurality of sample services according to the number of features of the target user feature, and use the service value of the target sample service as the centroid of each cluster corresponding to the target user feature;

[0037] an allocating unit, configured to allocate other sample services to the cluster where the corresponding centroid is located based on the similarity between the service values ​​of the other sample services and each centroid, wherein the other sample services are services other than the centroid in the sample services;

[0038] A first determining unit, configured to determine a service mean of each cluster;

[0039] A second determination unit is used for, when there is a target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, determining the service mean as the centroid corresponding to the target cluster, and returning to execute the step of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between other sample services and each centroid, until there is no target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, thereby obtaining the centroid of each cluster corresponding to the target user feature;

[0040] The conversion unit is used to convert the centroid of the cluster corresponding to each target user feature into a corresponding recommended service to obtain the recommended service corresponding to each target user feature.

[0041] In some embodiments, the dispensing unit comprises:

[0042] An acquisition subunit, used to acquire the similarity between the business values ​​of other sample businesses and each centroid;

[0043] The allocation subunit is used to allocate other sample services to the cluster where the target centroid with the smallest similarity to each centroid is located.

[0044] In some embodiments, obtaining a subunit includes:

[0045] An acquisition module, used for acquiring, for each centroid, a first sub-service value corresponding to each service type in other sample services, and a second sub-service value corresponding to each service type in the centroid;

[0046] A calculation module, used to calculate the difference between the first sub-service value and the second sub-service value corresponding to the same service type, and obtain the sub-service difference corresponding to each service type;

[0047] The determination module is used to determine the similarity between the business values ​​of other sample businesses and each centroid based on the sub-business difference corresponding to each business type.

[0048] In some embodiments, a determination module is used to:

[0049] Calculate the square of the sub-service difference corresponding to each service type to obtain a first calculation result;

[0050] Calculate the sum of multiple first calculation results to obtain a second calculation result;

[0051] The second calculation result is subjected to square root processing to obtain the similarity between the business values ​​of other sample businesses and each centroid.

[0052] In some embodiments, the first determining unit is configured to:

[0053] For each cluster, obtain the business volume of the services in the cluster and the business value corresponding to each service;

[0054] Calculate the sum of the service values ​​corresponding to each service to obtain a third calculation result;

[0055] A ratio of the third calculation result to the traffic volume is calculated to obtain a traffic mean of each cluster.

[0056] In some embodiments, the conversion unit is configured to:

[0057] Determine the target service value interval in which the centroid of each cluster corresponding to the target user feature is located;

[0058] The target service corresponding to each target service value interval is determined as a recommended service corresponding to the target user characteristics.

[0059] In some embodiments, the apparatus is further configured to:

[0060] Obtain the recommended user characteristics of the recommended user;

[0061] Filtering out a matching user feature that matches the user feature to be recommended from the plurality of target user features;

[0062] Recommend the recommended service corresponding to the matching user feature to the to-be-recommended user.

[0063] To achieve the above objectives, a 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 when the processor executes the computer program, the method described in the first aspect is implemented.

[0064] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0065] The service recommendation method, device, electronic device and storage medium proposed in the present application obtain sample services and user features of multiple users, determine the target user features from the multiple user features; select the target sample services from the multiple sample services according to the number of features of the target user features, and use the service value of the target sample services as the centroid of each cluster corresponding to the target user features; based on the similarity between the service values ​​of other sample services and each centroid, allocate the other sample services to the cluster where the corresponding centroid is located, and the other sample services are the services in the sample services other than the centroid; determine the Business mean; when there is a target cluster in multiple clusters where the difference between the centroid and the corresponding business mean reaches a preset difference value, the business mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample businesses to the cluster where the corresponding centroid is located based on the similarity between other sample businesses and each centroid is returned to execute, until there is no target cluster in multiple clusters where the difference between the centroid and the corresponding business mean reaches a preset difference value, and the centroid of each cluster corresponding to the target user feature is obtained; the centroid of each cluster corresponding to the target user feature is converted into a corresponding recommended business, and the recommended business corresponding to each target user feature is obtained. In this way, by clustering the sample businesses according to different user features in advance, different recommended businesses are determined for different user features according to the clustering results, so that the corresponding recommended business can be determined according to the user characteristics of the recommended user in the future, so as to make business recommendations, avoid multiple modifications to the insurance policy, and improve the efficiency of business recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 is a flow chart of a service recommendation method provided in an embodiment of the present application;

[0067] Figure 2 yes Figure 1 Flow chart of step S103 in FIG.

[0068] Figure 3 yes Figure 2 Flow chart of step S201 in FIG.

[0069] Figure 4 yes Figure 3 Flow chart of step S303 in FIG.

[0070] Figure 5 yes Figure 1 Flow chart of step S104 in FIG.

[0071] Figure 6 yes Figure 1 Flow chart of step S106 in FIG.

[0072] Figure 7 is another flow chart of the service recommendation method provided in the embodiment of the present application;

[0073] Figure 8 is a structural diagram of a business recommendation device provided in an embodiment of the present application;

[0074] Fig. 9 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0075] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0076] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0077] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0078] First, some nouns involved in this application are analyzed:

[0079] Net Promoter Score (NPS): An index that measures the likelihood that a customer will recommend a business or service to others. NPS questionnaires usually ask customers a simple question, such as "How likely are you to recommend our products / services to a friend or colleague?" Customers generally choose between 0 and 10 points.

[0080] A score of 0 means that they are not likely to recommend at all, and a score of 10 means that they are very likely to recommend. Based on the customers' answers, they are divided into three categories: Promoters (scoring 9-10 points): These customers are very satisfied with the product or service and are very likely to recommend it to others. Passives (scoring 7-8 points): They are satisfied with the product or service, but are unlikely to actively recommend it. Detractors (scoring 0-6 points): These customers are dissatisfied with the product or service and may spread negative reviews to others.

[0081] NPS is calculated by subtracting the percentage of detractors from the percentage of promoters. The NPS questionnaire can help companies understand customer satisfaction and loyalty to their products or services, so as to make targeted improvements and enhancements.

[0082] Based on this, the embodiments of the present application provide a business recommendation method, device, electronic device and storage medium, which aim to avoid multiple modifications to the insurance policy and improve the efficiency of business recommendation.

[0083] The service recommendation method, device, electronic device and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the service recommendation method in the embodiments of the present application is described.

[0084] The business recommendation method provided in the embodiment of the present application relates to the field of business processing technology. The business 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 a server cluster or distributed system composed of multiple physical servers, or 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 business recommendation method, etc., but is not limited to the above forms.

[0085] 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, etc. 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, etc. 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 through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0086] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these 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.

[0087] Figure 1 is an optional flowchart of the service recommendation method provided in the embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S106.

[0088] Step S101 : obtaining sample services and user characteristics of multiple users, and determining target user characteristics from the multiple user characteristics.

[0089] The user's sample business and basic information are collected through the NPS questionnaire. The NPS questionnaire includes questions such as the user's income, age, and the expected level of business handling corresponding to multiple business types. Among them, the income situation can be set in the form of an interval, the age can be set in the form of manual input, and the expected level of handling can be set to 11 scoring standards in [0,10]. The user scores the type of business they expect to handle to obtain the user's basic information.

[0090] Specifically, the user characteristics of the users are determined by performing data analysis on the basic information of multiple users. Taking the user characteristics used to characterize the risk preference of the users as an example, the specific analysis process is as follows:

[0091] Establish judgment rules: 1. Age factor: Generally speaking, young users may be more inclined to take risks because they have more time to make up for possible losses; while older users may be more conservative because their financial situation is relatively stable and they pay more attention to the preservation of assets. For example, a corresponding relationship between an age range and the degree of risk preference can be set, such as 20-30 years old may tend to have a higher risk preference, 30-50 years old may have a stable preference, and over 50 years old may have a conservative preference. 2. Income factor: High-income users usually have stronger financial strength to take risks and may be more inclined to aggressive investment strategies; low-income users may pay more attention to the security of funds and have a more conservative risk preference. The degree of risk preference can be divided according to the income range, and finally a comprehensive judgment of the user's risk preference can be made based on age factors and income factors.

[0092] For example, user A's basic information is 25 years old and monthly income is 3,000-5,000 yuan; user B's basic information is 30 years old and monthly income is 8,000-12,000 yuan; user C's basic information is 50 years old and monthly income is more than 20,000 yuan. For example, a low monthly income (such as 3,000-5,000 yuan) is conservative, a medium income (8,000-12,000 yuan) is stable, and a high income (more than 20,000 yuan) is aggressive. For user A (25 years old, monthly income of 3,000-5,000 yuan): From the perspective of age, 25 years old is at a relatively young stage and may have a certain risk tolerance, but combined with the situation of low monthly income, the overall judgment of his risk preference is conservative. For user B (30 years old, monthly income of 8,000-12,000 yuan): 30 years old is moderate in age, and the monthly income is also at a medium level. He is neither too young to take risks nor too high in income to pursue aggressive investment. The overall judgment is that his risk preference is stable. For user C (50 years old, monthly income of more than 20,000 yuan): 50 years old is relatively old and usually pays more attention to the stability of assets, but the user has a high monthly income and strong financial strength. He may be more willing to try some high-risk and high-return investments. Overall, he is judged to have an aggressive risk preference.

[0093] Specifically, after determining the user features corresponding to each user, multiple user features are classified and summarized to divide out multiple representative user features, and these representative user features are used as target user features.

[0094] For example, the user characteristics of user A are conservative risk preference, the user characteristics of user B are robust risk preference, the user characteristics of user C are aggressive risk preference, the user characteristics of user D are robust risk preference, and the user characteristics of user E are aggressive risk preference. We can conclude that there are three representative user characteristics, namely conservative risk preference, robust risk preference, and aggressive risk preference. These three user characteristics are then determined as target user characteristics.

[0095] Step S102 : selecting a target sample service from the plurality of sample services according to the number of features of the target user feature, and taking the service value of the target sample service as the centroid of each cluster corresponding to the target user feature.

[0096] After determining the target user features in step S101, target sample services having the same number of features as the target user features are randomly selected from multiple sample services, and the service values ​​of the target sample services are respectively used as the centroid of the cluster corresponding to each target user feature.

[0097] For example, the number of target user characteristics is three, and three sample businesses are selected from the sample businesses of many users as target sample businesses. For example, select user B (young, low-income, conservative risk preference, purchased basic medical insurance), user C (middle-aged, medium-income, stable risk preference, purchased a combination of life insurance and medical insurance), and user D (old, high-income, aggressive risk preference, purchased investment-type insurance) as target sample businesses. The business value of the target sample business is used as the centroid. The business value can be a quantitative representation of the insurance product portfolio. For example, a comprehensive value can be calculated based on factors such as the type, insured amount, and premium of the insurance product. Assuming that the business value of user B is 20, the business value of user C is 50, and the business value of user D is 80, then these three values ​​are respectively used as the centroids of the corresponding target user characteristics (young low-income conservative, middle-aged medium-income stable, and old high-income aggressive) clusters.

[0098] Step S103: Allocate other sample services to the cluster where the corresponding centroid is located based on the similarity between the service values ​​of other sample services and each centroid, wherein the other sample services are services other than the centroid in the sample services.

[0099] Among them, for other sample businesses (except the target sample business as the centroid), the similarity between their business values ​​and each centroid is calculated. The similarity can take into account multiple factors, such as age gap, income gap, risk preference gap, and differences in insurance product portfolios. Some distance measurement method can be used, such as Euclidean distance or cosine similarity.

[0100] For example, user E's business value is 45, age is middle-aged, income is medium, and risk preference is stable. Calculate the similarity between user E's business value and the three centroids respectively, and you may find that the similarity with the sample business corresponding to user C as the centroid is the highest. According to the similarity calculation results, user E is assigned to the cluster where user C is located. That is, it is considered that user E and user C have similar business needs.

[0101] Step S104: determine the service mean of each cluster.

[0102] For each cluster, the business mean of all sample services is calculated. Take the cluster where user C is located as an example. Assume that there are user C, user E and some other users in this cluster. In order to determine whether the sample service division is the optimal solution, it is necessary to calculate the business mean of each cluster, so that the business mean can be used to measure whether the sample service division is the optimal solution.

[0103] Step S105, when there is a target cluster among multiple clusters whose difference value between the centroid and the corresponding business mean reaches a preset difference value, the business mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample businesses to the cluster where the corresponding centroid is located based on the similarity between other sample businesses and each centroid is returned to be executed, until there is no target cluster among multiple clusters whose difference value between the centroid and the corresponding business mean reaches a preset difference value, and the centroid of each cluster corresponding to the target user feature is obtained.

[0104] Among them, if there is a target cluster among multiple clusters whose difference value between the centroid and the corresponding business mean reaches a preset difference value, it means that the similarity between the centroid in the target cluster and other sample businesses allocated to the target cluster is low, then the calculated business mean is used as the centroid corresponding to the target cluster, and the steps of allocating other sample businesses to the cluster where the corresponding centroid is located and subsequent steps based on the similarity between other sample businesses and each centroid are executed repeatedly until there is no target cluster among multiple clusters whose difference value between the centroid and the corresponding business mean reaches a preset difference value, indicating that at this time there is no situation where the business value of the sample business has a low similarity with the centroid of the cluster where it is located, and the centroid of the cluster corresponding to each target user feature is obtained.

[0105] For example, three representative users are selected as the initial centroids: user A, the centroid business value is 40, representing the initial centroid of the conservative cluster; user B, the centroid business value is assumed to be 60, representing the initial centroid of the robust cluster; user C, the centroid business value is assumed to be 80, representing the initial centroid of the aggressive cluster.

[0106] There are some other sample businesses: User D, business value is 38. After calculating the similarity with the three centroids, it is assigned to the conservative cluster where user A is. User E, business value is 62. It is assigned to the robust cluster where user B is. User F, business value is 82. It is assigned to the aggressive cluster where user C is. User G, business value is assumed to be 68, and is assigned to the robust cluster where user B is.

[0107] Calculate the difference between the business mean and centroid of each cluster: The conservative cluster now has users A and D, and the business mean is (40+38) / 2=39. The difference between the centroid and the business mean is |40-39|=1. The robust cluster now has users B, E, and G, and the business mean is (60+62+68) / 3=63.3. The difference between the centroid and the business mean is |60-61|=1. The aggressive cluster now has users C and F, and the business mean is (80+82) / 2=81. The difference between the centroid and the business mean is |80-81|=1.

[0108] The default difference value is 3. At this time, the robust cluster becomes the target cluster.

[0109] The business mean of the robust cluster, 63.3, is determined as the new centroid.

[0110] The similarities between other sample businesses and the three centroids (the centroid of the conservative cluster 40, the new robust cluster centroid 63.3, and the radical cluster centroid 80) are recalculated and assigned.

[0111] If user H was originally misjudged as aggressive, after recalculation, the business value is 65. It may have been assigned incorrectly before, but now it is assigned to the robust cluster after recalculation.

[0112] Calculate the business mean and difference again: Recalculate the business mean and centroid difference of each cluster: The business mean and centroid difference of the conservative cluster may remain unchanged or change slightly. The robust cluster now has users B, E, G, and H, and the business mean is recalculated. Assume it is 64. The difference between the centroid and the business mean is |63.3-64|=0.7. The business mean and centroid difference of the aggressive cluster may also change.

[0113] Repeat this process until there is no target cluster in multiple clusters where the difference between the centroid and the corresponding business mean reaches the preset difference value. Finally, the stable centroid of the cluster corresponding to each risk preference type is obtained, providing more accurate insurance recommendations for users with different risk preferences.

[0114] Step S106: convert the centroid of the cluster corresponding to each target user feature into a corresponding recommended service, and obtain the recommended service corresponding to each target user feature.

[0115] After the stable centroid of the cluster corresponding to each target user feature is determined in step S105, the value corresponding to the centroid can be converted into a corresponding recommended service to obtain the recommended service corresponding to each target user feature.

[0116] For example, for the centroid of the conservative cluster, we can recommend low-risk insurance products that guarantee basic living needs, such as small medical insurance and traditional life insurance; for the centroid of the steady cluster, we can recommend a portfolio of insurance products that balance risks and returns, such as a reasonable combination of life insurance, medical insurance, and accident insurance; for the centroid of the aggressive cluster, we can recommend high-risk and high-return investment insurance products, such as investment-linked insurance.

[0117] Steps S101 to S105 shown in the embodiment of the present application are to obtain sample services and user features of multiple users, determine target user features from the multiple user features; select target sample services from the multiple sample services according to the number of features of the target user features, and use the service value of the target sample service as the centroid of each cluster corresponding to the target user features; based on the similarity between the service value of other sample services and each centroid, assign other sample services to the cluster where the corresponding centroid is located, wherein the other sample services are services other than the centroid in the sample services; determine the service of each cluster Mean; when there is a target cluster in multiple clusters where the difference between the centroid and the corresponding business mean reaches a preset difference value, the business mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample businesses to the cluster where the corresponding centroid is located based on the similarity between other sample businesses and each centroid is returned to execute, until there is no target cluster in multiple clusters where the difference between the centroid and the corresponding business mean reaches a preset difference value, and the centroid of each cluster corresponding to the target user feature is obtained; the centroid of each cluster corresponding to the target user feature is converted into a corresponding recommended business, and the recommended business corresponding to each target user feature is obtained. In this way, by clustering the sample businesses according to different user features in advance, different recommended businesses are determined for different user features according to the clustering results, so that the corresponding recommended business can be determined according to the user characteristics of the recommended user in the future, so as to make business recommendations, avoid multiple modifications to the insurance policy, and improve the efficiency of business recommendations.

[0118] See also Figure 2 In some embodiments, step S103 may include but is not limited to steps S201 to S202:

[0119] Step S201, obtaining the similarity between the business values ​​of other sample businesses and each centroid.

[0120] For each other sample service, it is necessary to obtain the similarity between the service value of the other sample service and each centroid, so as to allocate the other sample service to a corresponding cluster among the multiple clusters according to the similarity.

[0121] Step S202: Allocate other sample services to the cluster where the target centroid with the smallest similarity to each centroid is located.

[0122] Among them, since the similarity represents the similarity of two businesses, the smaller the similarity, the closer the business values ​​of the two businesses are; the smaller the similarity, the farther the business values ​​of the two businesses are. Therefore, for each other sample business, each other sample business is assigned to the cluster where the target centroid with the smallest similarity is located among multiple centroids.

[0123] In this way, each other sample business is allocated by the similarity between its business value and each centroid. Accurate allocation of sample business helps improve the accuracy and personalization of recommendations. For insurance sellers, it can better meet user needs, improve user satisfaction and conversion rate; for users, they can find business products that suit them faster, saving time and energy.

[0124] See also Figure 3 In some embodiments, step S201 may include but is not limited to steps S301 to S303:

[0125] Step S301, for each centroid, obtain the first sub-service value corresponding to each service type in other sample services, and obtain the second sub-service value corresponding to each service type in the centroid.

[0126] Among them, since the sample services are generally presented in the form of a combination of multiple types of services, when calculating the similarity between each other sample service and each centroid, it is necessary to obtain the first sub-service value corresponding to each service type in the other sample services, and to obtain the second sub-service value corresponding to each service type in the centroid.

[0127] For example, there is a centroid representing the characteristics of a user with a stable risk preference, whose insurance business portfolio includes life insurance with a coverage of 400,000 yuan, medical insurance with a coverage of 200,000 yuan, and accident insurance with a coverage of 100,000 yuan, with an annual premium of 6,000 yuan; another sample business is the insurance situation of a user, including life insurance with a coverage of 380,000 yuan, medical insurance with a coverage of 180,000 yuan, and accident insurance with a coverage of 80,000 yuan, with an annual premium of 5,800 yuan; for this sample business and centroid, the first sub-business value of the life insurance business type is 380,000 yuan, and the second sub-business value is 400,000 yuan; the first sub-business value of the medical insurance business type is 180,000 yuan, and the second sub-business value is 200,000 yuan; the first sub-business value of the accident insurance business type is 80,000 yuan, and the second sub-business value is 100,000 yuan; the first sub-business value of the premium business type is 5,800 yuan, and the second sub-business value is 6,000 yuan.

[0128] Step S302, calculate the difference between the first sub-service value and the corresponding second sub-service value corresponding to the same service type, and obtain the sub-service difference corresponding to each service type.

[0129] Among them, by calculating the difference, we can intuitively understand the specific difference between other sample businesses and the centroid in each business type. The smaller the difference, the closer the two are in this business type, and vice versa. It provides a specific quantitative data basis for the subsequent determination of similarity, so as to measure the overall similarity by combining the differences of multiple business types.

[0130] Specifically, for each service type, the second sub-service value of the centroid is subtracted from the first sub-service value of other sample services to obtain the sub-service difference of the service type.

[0131] If the first sub-business value is greater than the second sub-business value, the difference is a positive number, indicating that the values ​​of other sample businesses on this business type are greater than the centroid value; if the first sub-business value is less than the second sub-business value, the difference is a negative number, indicating that the values ​​of other sample businesses on this business type are less than the centroid value.

[0132] For example, the sub-business difference of the life insurance business type is 380,000-400,000 = -20,000; the sub-business difference of the medical insurance business type is 180,000-200,000 = -20,000; the sub-business difference of the accident insurance business type is 80,000-100,000 = -20,000; the sub-business difference of the premium business type is 5,800 yuan-6,000 yuan = -200 yuan.

[0133] Step S303: Determine the similarity between the service values ​​of other sample services and each centroid based on the sub-service difference corresponding to each service type.

[0134] Among them, a quantitative similarity index is given by combining the differences of multiple business types to determine which cluster the centroid should be assigned to for other sample businesses. The higher the similarity, the more similar other sample businesses are to the centroid, and they should be assigned to the cluster where the centroid is located; the lower the similarity, the greater the difference between the two, and they should not be assigned to the cluster.

[0135] Specifically, a variety of methods can be used to determine the similarity, the most common of which are Euclidean distance, cosine similarity, etc. Taking Euclidean distance as an example, the squares of the sub-service differences of each service type are added together, and then the square root is taken. The result is the Euclidean distance between other sample services and the centroid. The smaller the distance, the higher the similarity.

[0136] The sub-service differences of different service types may also be weighted according to actual conditions, for example, a larger weight may be given to an important service type, and then the similarity may be calculated.

[0137] In this way, by calculating the similarity, users with similar business needs can be assigned to the same cluster, and then specific business recommendation plans can be formulated according to the characteristics of the users in this cluster to improve the pertinence and effectiveness of the recommendations.

[0138] See also Figure 4 In some embodiments, step S303 may include but is not limited to steps S401 to S403:

[0139] Step S401, calculate the square of the sub-service difference corresponding to each service type to obtain a first calculation result.

[0140] Among them, by squaring the difference, the impact of the difference can be magnified. Smaller differences will remain small after squaring, but larger differences will become larger after squaring, which can more clearly distinguish the differences between different business types. This provides an intermediate result that highlights the difference for the subsequent calculation of the overall similarity, so that the final similarity calculation more accurately reflects the actual difference between other sample businesses and the centroid.

[0141] For example, the sub-business difference of the life insurance business type is -20,000, which is squared to (-20,000). 2 =400,000,000; the difference of sub-business of medical insurance business type is -20,000, which is also 400,000,000 after square; the difference of sub-business of accident insurance business type is -20,000, which is also 400,000,000 after square; the difference of sub-business of premium business type is -200 yuan, which is (-200) after square 2 =40000.

[0142] Step S402, calculating the sum of multiple first calculation results to obtain a second calculation result.

[0143] The squared values ​​of the differences of each business type are summed up, and the differences of all business types are comprehensively considered to obtain a value reflecting the overall degree of difference. This provides an intermediate transition value for the final determination of the similarity between the sample business and the centroid. The larger this value is, the greater the comprehensive difference between the sample business and the centroid in multiple business types.

[0144] For example, the first calculation result of the life insurance business type is 400000000, the first calculation result of the medical insurance business type is 400000000, and the first calculation result of the premium business type is 40000, then the second calculation result is 400000000+400000000+40000=800040000.

[0145] Step S403: performing square root processing on the second calculation result to obtain the similarity between the business values ​​of other sample businesses and each centroid.

[0146] The square root processing is to convert the comprehensive difference value obtained in the previous step into a more practical similarity index. The result after the square root can balance the amplification effect of the difference to a certain extent, making the range of similarity values ​​more reasonable. The final result can intuitively reflect the similarity between the sample business and the centroid. The smaller the value, the higher the similarity, and the larger the value, the lower the similarity.

[0147] For example, if the second calculation result is 1200040000, the square root is about 34641.47. This result is the similarity between the sample business and the centroid.

[0148] Among them, for the sample business serving as the centroid, before calculating the similarity between each other sample business and the centroid, the time factor between other sample businesses and the sample business of the centroid can also be considered. For example, when calculating the similarity between a certain other sample business and the centroid, the business time difference between the other sample business and the sample business of the centroid is determined. The business time difference is the difference in the time of filling out the NPS questionnaire. The smaller the business time difference, the closer the weight value assigned to the sub-business value corresponding to each business type in the other sample business is to 1. In this way, when performing subsequent similarity calculations, the first sub-business value corresponding to each business type participating in the calculation in other sample businesses is the product of the original sub-business value and the weight, thereby improving the effectiveness of other sample businesses with smaller business time differences.

[0149] In this way, by squaring the sub-business difference of each business type, the difference between different business types can be effectively highlighted. In the insurance business, different business types may have different degrees of importance. The squaring operation can make larger differences more significant, thereby giving these differences greater weight in subsequent similarity calculations. An intermediate transition value is provided for the final determination of the similarity between the sample business and the centroid. This value can be used as a reference indicator to help the system determine the size of the comprehensive difference between the sample business and the centroid. In the insurance business, this intermediate transition value can be used to compare the degree of difference between different sample businesses and the centroid, so as to more reasonably allocate and recommend sample businesses. Taking the square root of the second calculation result can balance the amplification effect of the difference to a certain extent.

[0150] See also Figure 5 In some embodiments, step S104 may include but is not limited to steps S501 to S503:

[0151] Step S501 : for each cluster, obtain the service volume of services in the cluster and the service value corresponding to each service.

[0152] The business volume is used to determine the size of the business in the cluster, while the business value reflects the characteristics and importance of each specific business. This information is obtained in preparation for the subsequent calculation of the business mean, so as to better understand the overall characteristics and central trend of the cluster. By analyzing the number and value of the business in the cluster, the diversity and concentration of the cluster can be evaluated, providing a basis for further data analysis and decision-making.

[0153] Specifically, for each cluster, the business volume can be obtained by traversing all sample businesses in the cluster. The business volume can usually be obtained simply by counting, that is, counting the total number of sample businesses in the cluster. The business value corresponding to each business can be determined based on specific business characteristics. In insurance business, the business value can be the combined value of insurance products, the premium amount, the size of the insured amount, etc. The business value of each sample business can be read from a database or data storage structure.

[0154] For example, there are five insurance business samples in a cluster. The business volume is 5. For each business, assume that the business value is 1000 for business 1 (for example, representing the value of a certain insurance combination), 1200 for business 2, 900 for business 3, 1100 for business 4, and 1050 for business 5.

[0155] Step S502, calculating the sum of the service values ​​corresponding to each service to obtain a third calculation result.

[0156] Among them, calculating the sum of the business values ​​can obtain the overall value or total feature amount of all businesses in the cluster. This sum is an intermediate step in calculating the business mean, which is used to reflect the overall scale and feature accumulation of the businesses in the cluster. By summing up, the scattered business values ​​can be integrated to provide a total amount indicator for further calculation of the mean, so as to better understand the overall characteristics and central trend of the cluster.

[0157] Specifically, the service value corresponding to each service in the cluster is accumulated. Mathematical addition operation can be used to add all service values ​​one by one.

[0158] For example, the business value of business 1 is 1000, the business value of business 2 is 1200, the business value of business 3 is 900, the business value of business 4 is 1100, and the business value of business 5 is 1050. Add them together to get: 1000+1200+900+1100+1050=5250. This result is the third calculation result.

[0159] Step S503: Calculate the ratio of the third calculation result to the service volume to obtain the service mean of each cluster.

[0160] Among them, the business mean represents the average level of the business in the cluster, which can reflect the overall characteristics and concentration of the cluster. By calculating the business mean, we can better understand the representative characteristics of the business in the cluster and provide an important reference for subsequent data analysis and decision-making. For example, in the insurance business, the business mean can help determine the average insurance demand and value level of the users in the cluster, thereby providing a basis for formulating targeted insurance product recommendations and marketing strategies.

[0161] Specifically, the third calculation result is divided by the business volume to obtain the business mean, which can be calculated using mathematical division.

[0162] For example, the third calculation result is 5250, and the business volume is 5. Calculate the business mean: 5250 / 5=1050. This result is the business mean of the cluster, indicating that the average value level of insurance business in this cluster is 1050.

[0163] In this way, by calculating the service mean, we can better understand the representative characteristics of the services in the cluster.

[0164] See also Figure 6 In some embodiments, step S106 may include but is not limited to steps S601 to S602:

[0165] Step S601: determine the target service value interval where the centroid of each cluster corresponding to the target user feature is located.

[0166] After determining the centroid of the cluster corresponding to each target user feature, determining the service value interval where the centroid is located helps to further determine the service scope corresponding to different target user features. By clarifying the location of the centroid, accurate recommended services can be better provided to users with corresponding features.

[0167] Specifically, multiple business value intervals are set in advance. For combined insurance, business value intervals corresponding to different combinations can be set. For example, a combination insurance mainly based on basic medical insurance and a small amount of accident insurance has a relatively low insured amount, and the corresponding business value interval is 500-2000; a balanced combination of life insurance, medical insurance and accident insurance has a corresponding business value interval of 2500-8000; a combination of investment insurance and high-value medical insurance and life insurance has a corresponding business value interval of 8500-20000, etc., which is not limited here. Thus, the business value interval of each centroid is determined.

[0168] Step S602: Determine the target service corresponding to each target service value interval as a recommended service corresponding to the target user characteristics.

[0169] After the target service value interval in which the centroid of the cluster corresponding to each target user feature is located is determined in step S601, the recommended service corresponding to the target user feature can be determined according to the target service corresponding to the target service value interval.

[0170] In this way, by determining the business value range where the centroid is located, the insurance business scope corresponding to different target user characteristics can be clearly defined. In insurance recommendations, this helps to quickly understand the insurance combinations and value ranges that user groups with specific characteristics usually choose, providing an important reference for subsequent accurate recommendations.

[0171] See also Figure 7 In some embodiments, after step S106, the following steps may also be performed, including but not limited to steps S701 to S703:

[0172] Step S701, obtaining the recommended user characteristics of the recommended user.

[0173] Among them, we need to understand the specific situation of the user to be recommended so as to provide personalized recommendation services for him. User characteristics are important factors in determining the user's insurance needs and preferences. Obtaining these characteristics can better understand the user's needs and improve the accuracy and pertinence of the recommendation. This will prepare for the subsequent matching with the target user's characteristics and find the most suitable recommendation business for the user.

[0174] Specifically, the characteristics of the user to be recommended can be obtained in a variety of ways. For example, information can be collected when the user fills out a questionnaire, including age, gender, income level, family status, risk preference, etc. The characteristics of the user can also be inferred by analyzing the user's historical behavior data, such as purchased insurance products, browsing history, etc.

[0175] For example, suppose the user to be recommended is a 35-year-old male with a monthly income of 8,000 yuan, and these user characteristics are obtained through questionnaires or data analysis.

[0176] Step S702: Filter out matching user features that match the user features to be recommended from the multiple target user features.

[0177] Among them, by matching user characteristics, the recommended users can be classified into the most suitable user group, improving the accuracy of the recommendation, reducing the blindness of the recommendation, avoiding recommending insurance products that are not suitable for the characteristics of the users, and improving user satisfaction and acceptance.

[0178] Specifically, a variety of methods can be used for feature matching. A common method is to calculate the similarity between the features of the user to be recommended and the features of each target user. For example, a distance measurement method (such as Euclidean distance, cosine similarity, etc.) can be used to calculate the similarity. The higher the similarity, the closer the two user features are. According to the set similarity threshold or ranking, the target user features that best match the features of the user to be recommended are screened out.

[0179] For example, the target user characteristics include conservative risk preference, steady risk preference, aggressive risk preference, etc. If the risk preference of the to-be-recommended user characteristics is conservative, then the target user characteristics that best match the to-be-recommended user characteristics are determined to be conservative risk preference.

[0180] Step S703: recommending the recommended service corresponding to the matching user feature to the user to be recommended.

[0181] Among them, recommending corresponding businesses based on matching user characteristics can improve users' acceptance and satisfaction with recommendations and increase the sales conversion rate of insurance products. It can realize personalized recommendation services, meet users' specific needs, and improve user experience.

[0182] Specifically, the corresponding recommended services can be found based on the selected matching user characteristics. In the previous step, the corresponding recommended services have been determined for each target user characteristic. Services can be recommended to users in a variety of ways, such as displaying recommended products on the insurance sales platform, sending emails or SMS notifications, and having insurance sales staff make recommendations.

[0183] In this way, recommending corresponding services based on matching user characteristics can improve user acceptance and satisfaction with recommendations. In insurance recommendations, if the recommended products are consistent with the user's needs and preferences, the user will be more willing to accept and purchase these products. For example, if the user's risk preference is conservative, and the recommended insurance product portfolio also meets their risk tolerance and financial goals, the user will feel that these products are tailor-made for them, thereby improving satisfaction with the recommendation.

[0184] See also Figure 8 The embodiment of the present application further provides a service recommendation device, which can implement the above service recommendation method, and the device includes:

[0185] A first acquisition unit is used to acquire sample services and user features of multiple users, and determine target user features from the multiple user features;

[0186] A selection unit, configured to select a target sample service from the plurality of sample services according to the number of features of the target user feature, and use the service value of the target sample service as the centroid of each cluster corresponding to the target user feature;

[0187] an allocating unit, configured to allocate other sample services to the cluster where the corresponding centroid is located based on the similarity between the service values ​​of the other sample services and each centroid, wherein the other sample services are services other than the centroid in the sample services;

[0188] A first determining unit, configured to determine a service mean of each cluster;

[0189] A second determination unit is used for, when there is a target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, determining the service mean as the centroid corresponding to the target cluster, and returning to execute the step of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between other sample services and each centroid, until there is no target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, thereby obtaining the centroid of each cluster corresponding to the target user feature;

[0190] The conversion unit is used to convert the centroid of the cluster corresponding to each target user feature into a corresponding recommended service to obtain the recommended service corresponding to each target user feature.

[0191] In some embodiments, the dispensing unit comprises:

[0192] An acquisition subunit, used to acquire the similarity between the business values ​​of other sample businesses and each centroid;

[0193] The allocation subunit is used to allocate other sample services to the cluster where the target centroid with the smallest similarity to each centroid is located.

[0194] In some embodiments, obtaining a subunit includes:

[0195] An acquisition module, used for acquiring, for each centroid, a first sub-service value corresponding to each service type in other sample services, and a second sub-service value corresponding to each service type in the centroid;

[0196] A calculation module, used to calculate the difference between the first sub-service value and the second sub-service value corresponding to the same service type, and obtain the sub-service difference corresponding to each service type;

[0197] The determination module is used to determine the similarity between the business values ​​of other sample businesses and each centroid based on the sub-business difference corresponding to each business type.

[0198] In some embodiments, a determination module is used to:

[0199] Calculate the square of the sub-service difference corresponding to each service type to obtain a first calculation result;

[0200] Calculate the sum of multiple first calculation results to obtain a second calculation result;

[0201] The second calculation result is subjected to square root processing to obtain the similarity between the business values ​​of other sample businesses and each centroid.

[0202] In some embodiments, the first determining unit is configured to:

[0203] For each cluster, obtain the business volume of the services in the cluster and the business value corresponding to each service;

[0204] Calculate the sum of the service values ​​corresponding to each service to obtain a third calculation result;

[0205] A ratio of the third calculation result to the traffic volume is calculated to obtain a traffic mean of each cluster.

[0206] In some embodiments, the conversion unit is configured to:

[0207] Determine the target service value interval in which the centroid of each cluster corresponding to the target user feature is located;

[0208] The target service corresponding to each target service value interval is determined as a recommended service corresponding to the target user characteristics.

[0209] In some embodiments, the apparatus is further configured to:

[0210] Obtain the recommended user characteristics of the recommended user;

[0211] Filtering out a matching user feature that matches the user feature to be recommended from the plurality of target user features;

[0212] Recommend the recommended service corresponding to the matching user feature to the to-be-recommended user.

[0213] The service recommendation device provided in the embodiment of the present application obtains sample services and user features of multiple users through a first acquisition unit, and determines target user features from the multiple user features; the selection unit selects target sample services from the multiple sample services according to the number of features of the target user features, and respectively uses the service value of the target sample service as the centroid of each cluster corresponding to the target user features; the allocation unit allocates other sample services to the cluster where the corresponding centroid is located based on the similarity between the service value of other sample services and each centroid, and the other sample services are services other than the centroid in the sample services; the first determination unit determines the service value of each cluster. The second determination unit determines the business mean as the centroid corresponding to the target cluster when there is a target cluster in the multiple clusters where the difference between the centroid and the corresponding business mean reaches the preset difference value, and returns to execute the step of allocating other sample businesses to the cluster where the corresponding centroid is located based on the similarity between other sample businesses and each centroid, until there is no target cluster in the multiple clusters where the difference between the centroid and the corresponding business mean reaches the preset difference value, and obtains the centroid of each cluster corresponding to the target user feature; the conversion unit converts the centroid of each cluster corresponding to the target user feature into the corresponding recommended business, and obtains the recommended business corresponding to each target user feature. In this way, by clustering the sample businesses according to different user features in advance, different recommended businesses are determined for different user features according to the clustering results, so as to facilitate the subsequent determination of the corresponding recommended business according to the user features of the recommended user, thereby performing business recommendations, avoiding multiple modifications to the insurance policy, and improving the efficiency of business recommendations. Each other sample business is allocated by the similarity between the business value of each other sample business and each centroid, and accurate sample business allocation helps to improve the accuracy and personalization of recommendations. For insurance sellers, it can better meet the needs of users and improve user satisfaction and conversion rate; for users, they can find business products that suit them faster and save time and energy. By calculating the similarity, users with similar business needs can be assigned to the same cluster, and then a specific business recommendation plan can be formulated according to the characteristics of the cluster users to improve the pertinence and effectiveness of the recommendation. By squaring the sub-business difference of each business type, the difference between different business types can be effectively highlighted. In the insurance business, different business types may have different degrees of importance. The squaring operation can make larger differences more significant, so that these differences are given greater weight in the subsequent similarity calculation. An intermediate transition value is provided for the final determination of the similarity between the sample business and the centroid. This value can be used as a reference indicator to help the system judge the comprehensive difference between the sample business and the centroid. In the insurance business, this intermediate transition value can be used to compare the degree of difference between different sample businesses and the centroid, so as to more reasonably allocate and recommend sample businesses.Taking the square root of the second calculation result can balance the amplification effect of the difference to a certain extent. By calculating the business mean, the representative characteristics of the business in the cluster can be better understood. By determining the business value interval where the centroid is located, the insurance business scope corresponding to different target user characteristics can be clearly defined. In insurance recommendations, this helps to quickly understand the insurance portfolio and its value range that user groups with specific characteristics usually choose, providing an important reference for subsequent accurate recommendations. Recommending corresponding businesses based on matching user characteristics can improve user acceptance and satisfaction with the recommendations. In insurance recommendations, if the recommended products are consistent with the user's needs and preferences, the user will be more willing to accept and purchase these products. For example, if the user's risk preference is conservative, and the recommended insurance product portfolio also meets their risk tolerance and financial goals, the user will feel that these products are tailored for them, thereby improving satisfaction with the recommendation.

[0214] The specific implementation of the service recommendation device is substantially the same as the specific implementation of the service recommendation method described above, and will not be described in detail herein.

[0215] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above service recommendation method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0216] See also Fig. 9 , Fig. 9 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0217] The processor 901 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (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;

[0218] The memory 902 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 902 can store an operating system and other applications. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the service recommendation method of the embodiment of this application;

[0219] Input / output interface 903, used to implement information input and output;

[0220] Communication interface 904, used to realize communication interaction between the device and other devices, which can be realized by wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0221] A bus 905 that transmits information between the various components of the device (e.g., the processor 901, the memory 902, the input / output interface 903, and the communication interface 904);

[0222] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .

[0223] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned service recommendation method is implemented.

[0224] 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 disposed 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.

[0225] The service recommendation method, service recommendation device, electronic device and storage medium provided in the embodiments of the present application obtain sample services and user features of multiple users, determine target user features from the multiple user features; select target sample services from the multiple sample services according to the number of features of the target user features, and use the service value of the target sample service as the centroid of each cluster corresponding to the target user features; based on the similarity between the service value of other sample services and each centroid, allocate other sample services to the cluster where the corresponding centroid is located, wherein the other sample services are services other than the centroid in the sample services; determine each The business mean of each cluster; when there is a target cluster in which the difference between the centroid and the corresponding business mean reaches a preset difference value among multiple clusters, the business mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample businesses to the cluster where the corresponding centroid is located based on the similarity between other sample businesses and each centroid is returned to execute, until there is no target cluster in which the difference between the centroid and the corresponding business mean reaches a preset difference value among multiple clusters, and the centroid of each cluster corresponding to the target user feature is obtained; the centroid of each cluster corresponding to the target user feature is converted into a corresponding recommended business, and the recommended business corresponding to each target user feature is obtained. In this way, by clustering the sample businesses according to different user features in advance, different recommended businesses are determined for different user features according to the clustering results, so as to facilitate the subsequent determination of the corresponding recommended business according to the user features of the recommended user, thereby making business recommendations, avoiding multiple modifications to the insurance policy, and improving the efficiency of business recommendations.

[0226] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present 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 the present application are also applicable to similar technical problems.

[0227] Those skilled in the art will appreciate 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.

[0228] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0229] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0230] 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 sequence. 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 of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising 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.

[0231] It should be understood that in the present 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 objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single 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.

[0232] In the several embodiments provided in the present 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 only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, 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.

[0233] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0234] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0235] 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 is essentially 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, and the computer software product is stored in a storage medium, including multiple instructions to enable 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: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0236] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A business recommendation method, characterized in that: include: Acquire sample services and user characteristics of multiple users, and determine target user characteristics from the multiple user characteristics; According to the number of features of the target user features, a target sample service is selected from the plurality of sample services, and the service value of the target sample service is respectively used as the centroid of each cluster corresponding to the target user feature; Based on the similarity between the service values ​​of other sample services and each centroid, allocating other sample services to the cluster where the corresponding centroid is located, wherein the other sample services are services other than the centroid in the sample services; Determine the business mean of each cluster; When there is a target cluster in multiple clusters whose difference between the centroid and the corresponding service mean reaches a preset difference value, the service mean is determined as the centroid corresponding to the target cluster, and the step of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between other sample services and each centroid is returned to be executed, until there is no target cluster in multiple clusters whose difference between the centroid and the corresponding service mean reaches a preset difference value, and the centroid of each cluster corresponding to the target user feature is obtained; The centroid of the cluster corresponding to each target user feature is converted into a corresponding recommended service to obtain a recommended service corresponding to each target user feature.

2. The service recommendation method according to claim 1, characterized in that: The method of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between the service value of other sample services and each centroid includes: Obtain the similarity between the business values ​​of other sample businesses and each centroid; Assign other sample businesses to the cluster where the target centroid with the smallest similarity to each centroid is located.

3. The service recommendation method according to claim 2, characterized in that: The obtaining of the similarity between the service value of other sample services and each centroid includes: For each centroid, obtain the first sub-service value corresponding to each service type in other sample services, and obtain the second sub-service value corresponding to each service type in the centroid; Calculate the difference between the first sub-service value and the second sub-service value corresponding to the same service type to obtain the sub-service difference corresponding to each service type; Based on the sub-service difference corresponding to each service type, the similarity between the service values ​​of other sample services and each centroid is determined.

4. The service recommendation method according to claim 3, characterized in that: The determining the similarity between the service value of other sample services and each centroid based on the sub-service difference corresponding to each service type includes: Calculate the square of the sub-service difference corresponding to each service type to obtain a first calculation result; Calculate the sum of multiple first calculation results to obtain a second calculation result; The second calculation result is subjected to square root processing to obtain the similarity between the business values ​​of other sample businesses and each centroid.

5. The service recommendation method according to claim 1, characterized in that: The determining of the service mean of each cluster includes: For each cluster, obtain the business volume of the services in the cluster and the business value corresponding to each service; Calculate the sum of the service values ​​corresponding to each service to obtain a third calculation result; A ratio of the third calculation result to the traffic volume is calculated to obtain a traffic mean of each cluster.

6. The service recommendation method according to any one of claims 1 to 5, characterized in that: The converting the centroid of the cluster corresponding to each target user feature into a corresponding recommended service to obtain the recommended service corresponding to each target user feature includes: Determine the target service value interval in which the centroid of each cluster corresponding to the target user feature is located; The target service corresponding to each target service value interval is determined as a recommended service corresponding to the target user characteristics.

7. The service recommendation method according to any one of claims 1 to 5, characterized in that: After converting the centroid of the cluster corresponding to each target user feature into the corresponding recommended service to obtain the recommended service corresponding to each target user feature, the method further includes: Obtain the recommended user characteristics of the recommended user; Filtering out a matching user feature that matches the user feature to be recommended from the plurality of target user features; Recommend the recommended service corresponding to the matching user feature to the to-be-recommended user.

8. A business recommendation device, characterized in that: include: A first acquisition unit is used to acquire sample services and user features of multiple users, and determine target user features from the multiple user features; A selection unit, configured to select a target sample service from the plurality of sample services according to the number of features of the target user feature, and use the service value of the target sample service as the centroid of each cluster corresponding to the target user feature; an allocating unit, configured to allocate other sample services to the cluster where the corresponding centroid is located based on the similarity between the service values ​​of the other sample services and each centroid, wherein the other sample services are services other than the centroid in the sample services; A first determining unit, configured to determine a service mean of each cluster; A second determination unit is used for, when there is a target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, determining the service mean as the centroid corresponding to the target cluster, and returning to execute the step of allocating other sample services to the cluster where the corresponding centroid is located based on the similarity between other sample services and each centroid, until there is no target cluster in the plurality of clusters where the difference between the centroid and the corresponding service mean reaches a preset difference value, thereby obtaining the centroid of each cluster corresponding to the target user feature; The conversion unit is used to convert the centroid of the cluster corresponding to each target user feature into a corresponding recommended service to obtain the recommended service corresponding to each target user feature.

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 service recommendation 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 service recommendation method according to any one of claims 1 to 7 is implemented.

Images