Credit card product recommendation method and device, equipment, medium and product
By performing two clustering and optimization algorithms to select products, the problem that existing credit card product recommendation methods cannot achieve accurate recommendations, and more efficient credit card product recommendations are achieved.
Patent Information
- Application Number
- CN202411772528.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-04
- Publication Date
- 2025-04-01
AI Technical Summary
The existing credit card product recommendation methods cannot achieve accurate product recommendations to different customer groups, resulting in poor recommendation results.
By using the clustering algorithm to cluster twice based on the credit card behavior data and personal basic information data of the customer group to be recommended, the segmented customer group is obtained; then, based on the characteristics of each customer group, an optimization algorithm is used to select the target recommended product and send it to the corresponding customer group.
It realizes accurate credit card product recommendations to different customer groups, improving the accuracy and effectiveness of recommendations.
Smart Images
Figure CN119741083A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of financial technology or other related fields, and in particular to a credit card product recommendation method, device, equipment, medium and product. Background Art
[0002] In an environment of rapid social and economic development, credit cards have become the first choice for many users due to their security and convenience; therefore, it is necessary to recommend corresponding credit card products for different customer groups.
[0003] At present, the recommendation method for some credit card products is usually that banks hold various promotional activities to make recommendations. However, this method cannot achieve accurate product recommendations for different customer groups. Summary of the invention
[0004] The present application provides a credit card product recommendation method, device, equipment, medium and product to improve the accuracy of credit card product recommendations.
[0005] In a first aspect, the present application provides a credit card product recommendation method, comprising: based on first historical information of a customer group to be recommended, clustering the customer group to be recommended using a clustering algorithm to obtain a first customer group; the first historical information includes credit card user behavior data of the customer group to be recommended; based on second historical information of the first customer group, clustering the first customer group using a clustering algorithm to obtain multiple second customer groups; the second historical information includes personal basic information data and financial information data of the first customer group; based on the products to be recommended corresponding to each second customer group, using an optimization algorithm to obtain target recommended products for each second customer group; and sending the target recommended products to the corresponding second customer groups.
[0006] In a second aspect, the present application provides a credit card product recommendation device, comprising: a classification module, used to cluster the customer group to be recommended by using a clustering algorithm to obtain a first customer group based on first historical information of the customer group to be recommended; the first historical information includes credit card user behavior data of the customer group to be recommended; the classification module is also used to cluster the first customer group by using a clustering algorithm to obtain multiple second customer groups based on second historical information of the first customer group; the second historical information includes personal basic information data and financial information data of the first customer group; a recommendation module is used to obtain target recommended products for each second customer group based on the products to be recommended corresponding to each second customer group by using an optimization algorithm; and the target recommended products are sent to the corresponding second customer groups.
[0007] In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the above method.
[0008] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the above method.
[0009] In a fifth aspect, the present application provides a computer program product, including a computer program, which is used to implement the above method when executed by a processor.
[0010] In the credit card product recommendation method, device, equipment, medium and product provided by the present application, a first customer group is obtained by clustering the credit card behavior data of the customer group to be recommended using a clustering algorithm; then, based on the obtained personal basic information data and financial data of the first customer group, a clustering algorithm is used again to cluster to obtain multiple second customer groups; then, based on the corresponding products to be recommended of each second customer group, an optimization algorithm is used to obtain the target recommended products corresponding to each second customer group. The scheme of the present application makes the classification of the customer group to be recommended more accurate by clustering the customer group to be recommended twice using a clustering algorithm; further, based on the products to be recommended corresponding to the customer group, an optimization algorithm is used to obtain the target recommended products among the products to be recommended, and send them to the corresponding customer group; thus, accurate recommendation of credit card products is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0012] Figure 1 A flowchart of a credit card product recommendation method is shown in FIG.
[0013] Figure 2 A flowchart of a credit card product recommendation method is shown in FIG.
[0014] Figure 3 A flowchart of a credit card product recommendation method is shown in FIG.
[0015] Figure 4 A flowchart of a credit card product recommendation method is shown in FIG.
[0016] Figure 5 A structural schematic diagram of a credit card product recommendation device is exemplarily shown in FIG.
[0017] Figure 6 A structural schematic diagram of an electronic device is exemplarily shown in FIG.
[0018] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION
[0019] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.
[0020] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.
[0021] It should be noted that the brief description of the terms in this application is only for the convenience of understanding the implementation methods described below, and is not intended to limit the implementation methods of this application. Unless otherwise specified, these terms should be understood according to their ordinary and usual meanings. The terms "first", "second", etc. in the specification and claims and the above-mentioned drawings in this application are used to distinguish similar or similar objects or entities, and do not necessarily mean to limit a specific order or sequence, unless otherwise indicated. It should be understood that the terms used in this way can be interchangeable where appropriate, for example, they can be implemented in an order other than those given in the diagram or description of the embodiment of this application. The terms "including" and "having" in the specification and claims and the above-mentioned drawings in this application and any of their variations are intended to cover but not exclude inclusion, for example, a product or device containing a series of components is not necessarily limited to those components clearly listed, but may include other components that are not clearly listed or inherent to these products or devices. The term "module" used in this application refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic or a combination of hardware or / and software code that can perform functions related to the element.
[0022] It should be noted that the credit card product recommendation method, device, equipment, medium and product provided in this application can be used in the field of financial technology, and can also be used in any field outside of financial technology. The application field of the credit card product recommendation method, device, equipment, medium and product in this application is not limited.
[0023] In an environment of rapid social and economic development, credit cards have become the first choice for many users due to their security and convenience. Therefore, it is necessary to recommend corresponding credit card products for different customer groups.
[0024] At present, the recommendation method for some credit card products is usually that banks hold various promotional activities to recommend products. However, due to different customer groups, customer needs are also different; this method cannot achieve accurate product recommendations for different customer groups.
[0025] The technical content provided by this application is intended to solve the above technical problems of the prior art. In the credit card product recommendation method, device, equipment, medium and product of this application, a first customer group is obtained by clustering the credit card behavior data of the customer group to be recommended using a clustering algorithm; then, based on the obtained personal basic information data and financial data of the first customer group, a clustering algorithm is used again to cluster to obtain multiple second customer groups; then, based on the corresponding products to be recommended of each second customer group, an optimization algorithm is used to obtain the target recommended products corresponding to each second customer group. The solution of this application uses a clustering algorithm to perform two clusterings on the customer group to be recommended, so that the classification of the customer group to be recommended is more accurate; further, based on the products to be recommended corresponding to the customer group, an optimization algorithm is used to obtain the target recommended products among the products to be recommended, and send them to the corresponding customer group; thus, accurate recommendation of credit card products is achieved.
[0026] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.
[0027] Embodiment 1
[0028] Figure 1 A flow chart of a credit card product recommendation method is shown in FIG. Figure 1 As shown, the method includes:
[0029] Step 101: cluster the customer group to be recommended using a clustering algorithm to obtain a first customer group according to first historical information of the customer group to be recommended; the first historical information includes credit card user behavior data of the customer group to be recommended;
[0030] Step 102: cluster the first customer group using a clustering algorithm to obtain a plurality of second customer groups according to the second historical information of the first customer group; the second historical information includes basic personal information data and financial information data of the first customer group;
[0031] Step 103: According to the to-be-recommended products corresponding to each second customer group, an optimization algorithm is used to obtain target recommended products for each second customer group; and the target recommended products are sent to the corresponding second customer group.
[0032] Specifically, according to the first historical information of the customer to be recommended, a clustering algorithm is used to cluster the customer group to be recommended to obtain a first customer group. It should be noted that this example does not limit the specific type of clustering algorithm. The first historical information includes the user card behavior data of the customer group to be recommended; in actual applications, the credit card behavior data of the user usually refers to various types of transaction and consumption data generated by the user in the process of using the credit card, as well as the credit records related thereto; including consumer finance related data, such as consumption amount, consumption times, consumption categories, etc.; the user's cash withdrawal amount, cash withdrawal times, repayment amount, repayment date and overdue amount, etc.; the user's credit card credit limit and monthly credit limit utilization rate. At the same time, it also includes some relevant data of users participating in credit card activities; such as the actual number of participation in credit card activities, the number of times the relevant activities are shared with others, and the number of conversions converted to bank customers or other goals through relevant credit card activities. Based on the above-mentioned relevant data of the user to be recommended, the customer group to be recommended is clustered using a clustering algorithm to obtain a first customer group; according to the second historical information data and financial information data of the first customer group, a clustering algorithm is used again to cluster to obtain multiple second customer groups. Exemplarily, the basic personal information data includes the age, gender, income level, educational background and other information of the recommended customer; the financial information data includes personal asset information, frequency of large-amount consumption, rate of return on bank products and other information. After obtaining multiple second customer groups, the target recommended products are selected from the recommended products corresponding to each second customer group using an optimization algorithm; in actual applications, the use of an optimization algorithm to select a target product from multiple products usually refers to evaluating and comparing a series of alternative products to find the best product that meets specific goals or standards. This example does not limit the specific type of optimization algorithm. After the target recommended product is found using the optimization algorithm, the target recommended product is recommended to the corresponding second customer group.
[0033] In this example, a two-step clustering algorithm is used to accurately classify the customer groups to be recommended. Afterwards, an optimization algorithm is used to find the right target products from the products to be recommended and recommend them to the corresponding customer groups, thereby improving the accuracy of credit card product recommendations.
[0034] Optionally, it is necessary to obtain information data corresponding to the user to be recommended; accordingly, as an example, Figure 2 A flowchart of a credit card product recommendation method is exemplified; based on any example, according to the first historical information of the customer group to be recommended, the customer group to be recommended is clustered using a clustering algorithm to obtain a first customer group, including:
[0035] Step 201: Based on the first historical information, the customer groups to be recommended are classified by using a clustering algorithm to obtain a plurality of customer groups of a preset number;
[0036] Step 202: Select a customer group whose credit card behavior data is higher than a preset value from multiple customer groups as the first customer group.
[0037] Specifically, according to the first historical information of the customer to be recommended, a clustering algorithm is used to cluster the customer group to be recommended to obtain a preset number of cluster groups; the number of cluster groups can be determined according to specific business needs and goals and the dimensions of the customer data to be recommended; in this example, after the customer group to be recommended is clustered into a set of multiple customer groups according to the specific number of clusters, the customer group whose customer credit card behavior data is higher than the preset value in the multiple customer groups is taken as the first customer group of this example. Exemplarily, according to the credit card behavior data of the user, a clustering algorithm is used to divide the customer into four customer groups, namely, a high communication value customer group, a medium communication value customer group, a low communication value customer group and a negative communication value customer group; wherein, the high communication value customer group indicates that the customer's credit card behavior data is at the first level; the medium communication value customer group indicates that the customer's credit card behavior data is at the second level; the low communication value customer group indicates that the customer's credit card behavior data is at the third level; the negative communication value customer group indicates that the credit card behavior data is at the fourth level; the customer groups corresponding to the first level and the second level are selected as a customer group of this example, and credit card products are recommended for the first customer group later. In this example, a preset value of the credit card behavior data is preset, and the corresponding customer group whose credit card behavior data is higher than the preset value is used as the first customer group in this example, thereby improving the accuracy of classification of the recommended customer group.
[0038] Furthermore, the credit card behavior data of the users to be recommended is required to select the first customer group; accordingly, as an example, based on any example, step 201 specifically includes:
[0039] Obtaining first historical information;
[0040] According to the target attributes, feature extraction is performed on the first historical information to obtain feature data; the target attributes include the number of credit card participations, the number of shares, and the number of conversions;
[0041] According to the characteristic data, a clustering algorithm is used to perform clustering processing to obtain a plurality of preset numbers of customer groups.
[0042] Specifically, first, the first historical information needs to be obtained; this example does not limit the method of obtaining the first historical information of the customer to be recommended, for example, it can be directly obtained from the database, or it can be obtained using other data acquisition methods. Then, the actual business needs and the target attributes of the customer group to be recommended are set for clustering, and the first historical information is feature extracted according to the target attributes to extract feature data; the target attributes of this example are the number of credit card participations, credit card sharing times, and credit card conversions of the customers in the customer group to be recommended; in actual applications, the number of credit card participations refers to the frequency of customers participating in credit card related activities. The number of credit card sharing times refers to the number of times a credit card user shares credit card related information, preferential activities, or usage experience with others; this indicator can reflect the word-of-mouth communication and user satisfaction of credit cards. The number of credit card conversions usually refers to the number of people who have changed from non-credit card users to credit card users, or the number of people who have changed from non-specific credit card users of a certain bank to specific credit card users; this indicator can reflect the market development capabilities and user appeal of credit cards.
[0043] Furthermore, based on the extracted feature data, a clustering algorithm is used to perform clustering processing to obtain a plurality of preset number of customer groups. In this example, by classifying the recommended customers based on the feature data extracted from the target attribute, the customer groups obtained by clustering are more focused on the target attribute, and credit card products are subsequently recommended to such customer groups, making the recommended products more accurate.
[0044] Exemplarily, it is necessary to use a clustering algorithm to obtain multiple cluster groups according to specific steps; accordingly, as an example, Figure 3 A flowchart of a credit card product recommendation method is exemplified; based on any example, a clustering algorithm is used to perform clustering processing to obtain a plurality of preset customer groups, including:
[0045] Step 301, determine the number of cluster groups;
[0046] Step 302: randomly select the first cluster center value of each cluster group;
[0047] Step 303: obtaining updated cluster groups based on the distance function according to the first cluster center values and the cluster objects in each cluster group; the cluster objects represent each customer object in the customers to be recommended; the distance function is used to define the proximity between each cluster object and each first cluster center value;
[0048] Step 304: Calculate the mean of each updated cluster group, and use each mean as the second cluster center value of each updated cluster group;
[0049] Step 305: Use the second cluster center value as the cluster center value of the next iteration, repeat the iterative clustering until the cluster center value of each cluster group reaches stability, then stop the iteration, and obtain a plurality of preset numbers of customer groups.
[0050] First, it is necessary to determine the number of clusters, that is, to determine how many customer groups the customer group to be recommended should be divided into; the number of clusters is determined according to the aforementioned specific business needs and goals and the dimensions of the customer data to be recommended. Then select the first cluster center value of each cluster group; as an example, for each cluster group, a point can be selected as the initial cluster center value by random selection; as another example, a point can be randomly selected as the initial cluster center of the first cluster group, and then the point farthest from the point is selected as the initial cluster center of the second cluster group; the point farthest from the selected center point of the previous cluster group is selected as the initial cluster center of the next cluster group in turn, until the initial cluster centers of all cluster groups are selected. Next, multiple updated cluster groups are obtained according to the distance function; wherein the distance function represents the degree of proximity between each cluster object and each first cluster center value. After obtaining multiple updated cluster groups, the mean of the cluster objects in each cluster group is calculated, and the mean corresponding to each cluster group is used as the center value of the next clustering; the iteration is repeated until the cluster center value of each cluster group reaches stability, and then the iteration is stopped; as an example, when the iteration reaches the preset number of iterations, the iteration is stopped. After the iteration is completed, multiple preset numbers of customer groups can be obtained. In this example, the clustering algorithm is used to cluster the recommended customer groups, which improves the accuracy of customer group classification.
[0051] Further, according to each first cluster center value and the cluster objects in each cluster group, each updated cluster group is obtained based on a distance function, including:
[0052] Calculate the distance between each cluster object and each first cluster center value, and select the first cluster center value with the shortest distance to each cluster object as the first cluster center value of the cluster group where the cluster object belongs;
[0053] The clustering object under the cluster group where the first cluster center value is located is determined according to each first cluster center value to obtain an updated cluster group.
[0054] According to the first cluster center value of each cluster group and each cluster object, the distance between each cluster object (that is, each data object) in the entire customer group to be recommended and the initial first cluster center value corresponding to each cluster group is calculated; the first cluster center with the smallest distance is selected as the center value of the data object; according to each first cluster center value, the cluster object under the cluster group where the first cluster center value is located is determined to obtain the updated cluster group. In this example, by calculating the distance between each cluster object and each center value, the center value corresponding to the cluster object with the smallest distance to it is selected as the new center value, which further improves the accuracy of the classification of the customer group to be recommended.
[0055] After clustering the recommended customer group to obtain multiple cluster sets, it is necessary to recommend credit card products to the recommended customer group; accordingly, as an example, Figure 4 A flowchart of a credit card product recommendation method is exemplified; based on any example, an optimization algorithm is used to obtain a target recommended product for each second customer group, including:
[0056] Step 401, calculating the fitness of each product to be recommended and each second customer group; the fitness represents the matching degree of each product to be recommended and each second customer group;
[0057] Step 402: According to the fitness, select a target recommended product based on the product to be recommended.
[0058] Specifically, first set a fitness evaluation criterion, which can be formulated according to actual business needs; calculate the fitness of each product to be recommended and each second customer group; where the fitness represents the degree of match between each product to be recommended and the second customer group; the higher the fitness value, the higher the degree of match between the product to be recommended and the evaluation criteria; select the target recommended product from the products to be recommended according to the fitness. In one example, a particle swarm algorithm is used to select the target recommended product; in actual applications, the particle swarm algorithm is an intelligent optimization search method based on the group; the particle swarm optimization algorithm searches the solution space by continuously updating the speed and position of the particles, and each particle updates its speed and position based on its own experience and the experience of the group; specifically, each particle calculates the new speed and position based on the current speed and position and its own historical optimal position and the historical optimal position of the group. Through continuous iterative updates, the particle swarm gradually approaches the global optimal solution. The usual steps of using the particle swarm optimization algorithm are: first initialize the population, then calculate the fitness of each object in the population (the product to be recommended in this application); update the particle speed and particle position according to the fitness; calculate the fitness at the new position, if the fitness of the new position is higher, then update the position of the particle, otherwise do not update; repeat the iteration until the termination condition is met; then output the optimal value, which is the target recommended product of this application. This example uses the optimization algorithm to select the target recommended product according to the fitness, so that the recommendation results of credit card products are more accurate.
[0059] Furthermore, according to the fitness, target recommended products are selected based on the products to be recommended, including:
[0060] According to the fitness, the recommended products with fitness higher than the preset value are selected as the target recommended products.
[0061] In actual applications, based on the fitness of each to-be-recommended product and each second customer group, the to-be-recommended product with a fitness higher than a preset value is selected as the target recommended product; the target recommended product is subsequently recommended to the corresponding second customer group. As an example, there may be multiple target recommended products, and multiple target recommended products are recommended to the user, and the customer can subsequently select products from the multiple target recommended products. As another example, multiple target recommended products can be sorted according to fitness, and credit card products can be recommended to the recommended customers according to priority based on the sorting results. In this example, the target recommended product corresponding to the second customer group is selected according to the optimization algorithm and recommended to the second customer group, thereby improving the accuracy of credit card product recommendations.
[0062] In one example, when it is necessary to recommend credit card products to customers, the customer group to be recommended is first clustered into four cluster groups using a clustering algorithm, that is, the customer group to be recommended is divided into four customer groups according to the user behavior data of the customers to be recommended; the four customer groups are: a high communication value customer group, a medium communication value customer group, a low communication value customer group, and a negative communication value customer group; among them, the high communication value customer group indicates that these customers often participate in bank activities and share the activities with other customers, and the shared customers also participate in bank activities, and the activity conversion rate is high; the medium communication value customer group indicates that these customers occasionally participate in bank activities, but as long as they participate, they will share the activities with other customers, and the shared customers also participate in bank activities, and the activity conversion rate is relatively high; the low communication value customer group indicates that these customers occasionally participate in bank activities, and the frequency of sharing activities is also very low, and the activity conversion rate is low; the negative communication value customer group indicates that these customers often participate in bank activities, but the frequency of sharing activities is very low, and the activity conversion rate is very low.
[0063] The high communication value group and the medium communication value group are selected as the first customer group; then, according to the personal basic information data and financial information data of the aforementioned first customer group, a clustering algorithm is used for clustering to obtain four cluster groups, namely, the first customer group is divided into four customer groups; the four customer groups are: high net worth high potential customer group, high net worth low potential customer group, low net worth high potential customer group, low net worth low potential customer group; among them, the high net worth high potential customer group indicates that these customers already have high value in the bank and have great potential for higher value in the future; the high net worth low potential customer group indicates that these customers already have high value in the bank, but have little potential for higher value in the future; the low net worth high potential customer group indicates that these customers do not have high value in the bank, but have great potential for higher value in the future; the low net worth high potential customer group indicates that these customers do not have high value in the bank and do not have great potential for higher value in the future.
[0064] Next, credit card products are recommended based on multiple credit card products to be recommended. Exemplarily, the attributes corresponding to the credit card products to be recommended are the credit card consumption amount and the number of card swipes; for high-net-worth and high-potential customer groups, the recommended credit card products correspond to preferential plans with a total of 1,000-400; for high-net-worth and low-potential customer groups, the recommended credit card products correspond to preferential plans with a total of 1,000-300; for low-net-worth and high-potential customer groups, the recommended credit card products correspond to a plan for giving a 10-yuan discount for swiping the credit card 10 times; for low-net-worth and low-potential customer groups, the recommended credit card products correspond to a plan for giving a 10-yuan discount for swiping the card 15 times and the consumption amount reaching 200 yuan. Use an optimization algorithm to find a credit card product with a fitness higher than a preset value as the target recommended product for multiple credit card products to be recommended; wherein, the target recommended product can be one or more.
[0065] In the credit card product recommendation method provided in this embodiment, a clustering algorithm is used to cluster the credit card behavior data of the customer group to be recommended to obtain a first customer group; then, a clustering algorithm is used again to cluster the personal basic information data and financial data of the first customer group to obtain multiple second customer groups; then, an optimization algorithm is used to obtain the target recommended products corresponding to each second customer group based on the corresponding products to be recommended of each second customer group. The solution of this application accurately classifies the customer group to be recommended by clustering the customer group to be recommended twice, and then recommends products to the second customer group based on multiple products to be recommended using an optimization algorithm; the solution of this application improves the accuracy of credit card product recommendations.
[0066] Embodiment 2
[0067] Figure 5 The structure diagram of a credit card product recommendation device is shown as an example; Figure 5 As shown, the device comprises:
[0068] The classification module 21 is used to cluster the customer group to be recommended using a clustering algorithm to obtain a first customer group according to the first historical information of the customer group to be recommended; the first historical information includes the credit card user behavior data of the customer group to be recommended;
[0069] The classification module 21 is further used to cluster the first customer group using a clustering algorithm to obtain a plurality of second customer groups according to the second historical information of the first customer group; the second historical information includes basic personal information data and financial information data of the first customer group;
[0070] The recommendation module 22 is used to obtain target recommended products for each second customer group using an optimization algorithm based on the products to be recommended corresponding to each second customer group; and send the target recommended products to the corresponding second customer group.
[0071] Specifically, according to the first historical information of the customer to be recommended, a clustering algorithm is used to cluster the customer group to be recommended to obtain a first customer group. It should be noted that this example does not limit the specific type of clustering algorithm. The first historical information includes the user card behavior data of the customer group to be recommended; in actual applications, the credit card behavior data of the user usually refers to various types of transaction and consumption data generated by the user in the process of using the credit card, as well as the credit records related thereto; including consumer finance related data, such as consumption amount, consumption times, consumption categories, etc.; the user's cash withdrawal amount, cash withdrawal times, repayment amount, repayment date and overdue amount, etc.; the user's credit card credit limit and monthly credit limit utilization rate. At the same time, it also includes some relevant data of users participating in credit card activities; such as the actual number of participation in credit card activities, the number of times the relevant activities are shared with others, and the number of conversions converted to bank customers or other goals through relevant credit card activities. Based on the above-mentioned relevant data of the user to be recommended, the customer group to be recommended is clustered using a clustering algorithm to obtain a first customer group; according to the second historical information data and financial information data of the first customer group, a clustering algorithm is used again to cluster to obtain multiple second customer groups. Exemplarily, the basic personal information data includes the age, gender, income level, educational background and other information of the recommended customer; the financial information data includes personal asset information, frequency of large-amount consumption, rate of return on bank products and other information. After obtaining multiple second customer groups, the target recommended products are selected from the recommended products corresponding to each second customer group using an optimization algorithm; in actual applications, the use of an optimization algorithm to select a target product from multiple products usually refers to evaluating and comparing a series of alternative products to find the best product that meets specific goals or standards. This example does not limit the specific type of optimization algorithm. After the target recommended product is found using the optimization algorithm, the target recommended product is recommended to the corresponding second customer group.
[0072] In this example, a two-step clustering algorithm is used to accurately classify the customer groups to be recommended. Afterwards, an optimization algorithm is used to find the right target products from the products to be recommended and recommend them to the corresponding customer groups, thereby improving the accuracy of credit card product recommendations.
[0073] Optionally, it is necessary to obtain information data corresponding to the user to be recommended; accordingly, as an example, the classification module 21 is used to:
[0074] According to the first historical information, the customer groups to be recommended are classified by using a clustering algorithm to obtain a plurality of customer groups of a preset number;
[0075] A customer group whose credit card behavior data is higher than a preset value is selected from multiple customer groups as the first customer group.
[0076] Specifically, according to the first historical information of the customer to be recommended, a clustering algorithm is used to cluster the customer group to be recommended to obtain a preset number of cluster groups; the number of cluster groups can be determined according to specific business needs and goals and the dimensions of the customer data to be recommended; in this example, after the customer group to be recommended is clustered into a set of multiple customer groups according to the specific number of clusters, the customer group whose customer credit card behavior data is higher than the preset value in the multiple customer groups is taken as the first customer group of this example. Exemplarily, according to the credit card behavior data of the user, a clustering algorithm is used to divide the customer into four customer groups, namely, a high communication value customer group, a medium communication value customer group, a low communication value customer group and a negative communication value customer group; wherein, the high communication value customer group indicates that the customer's credit card behavior data is at the first level; the medium communication value customer group indicates that the customer's credit card behavior data is at the second level; the low communication value customer group indicates that the customer's credit card behavior data is at the third level; the negative communication value customer group indicates that the credit card behavior data is at the fourth level; the customer groups corresponding to the first level and the second level are selected as a customer group of this example, and credit card products are recommended for the first customer group later. In this example, a preset value of the credit card behavior data is preset, and the corresponding customer group whose credit card behavior data is higher than the preset value is used as the first customer group in this example, thereby improving the accuracy of classification of the recommended customer group.
[0077] Furthermore, the credit card behavior data of the users to be recommended is needed to select the first customer group; accordingly, as an example, the classification module 21 is specifically used for:
[0078] Obtaining first historical information;
[0079] According to the target attributes, feature extraction is performed on the first historical information to obtain feature data; the target attributes include the number of credit card participations, the number of shares, and the number of conversions;
[0080] According to the characteristic data, a clustering algorithm is used to perform clustering processing to obtain a plurality of preset numbers of customer groups.
[0081] Specifically, first, the first historical information needs to be obtained; this example does not limit the method of obtaining the first historical information of the customer to be recommended, for example, it can be directly obtained from the database, or it can be obtained using other data acquisition methods. Then, the actual business needs and the target attributes of the customer group to be recommended are set for clustering, and the first historical information is feature extracted according to the target attributes to extract feature data; the target attributes of this example are the number of credit card participations, credit card sharing times, and credit card conversions of the customers in the customer group to be recommended; in actual applications, the number of credit card participations refers to the frequency of customers participating in credit card related activities. The number of credit card sharing times refers to the number of times a credit card user shares credit card related information, preferential activities, or usage experience with others; this indicator can reflect the word-of-mouth communication and user satisfaction of credit cards. The number of credit card conversions usually refers to the number of people who have changed from non-credit card users to credit card users, or the number of people who have changed from non-specific credit card users of a certain bank to specific credit card users; this indicator can reflect the market development capabilities and user appeal of credit cards.
[0082] Furthermore, based on the extracted feature data, a clustering algorithm is used to perform clustering processing to obtain a plurality of preset number of customer groups. In this example, by classifying the recommended customers based on the feature data extracted from the target attribute, the customer groups obtained by clustering are more focused on the target attribute, and credit card products are subsequently recommended to such customer groups, making the recommended credit card products more accurate.
[0083] Exemplarily, it is necessary to use a clustering algorithm to obtain multiple cluster groups according to specific steps; accordingly, as an example, a clustering algorithm is used to perform clustering processing to obtain multiple preset numbers of customer groups, including:
[0084] Determine the number of clusters;
[0085] Randomly select the first cluster center value of each cluster group;
[0086] According to each first cluster center value and the cluster objects in each cluster group, each updated cluster group is obtained based on the distance function; the cluster object represents each customer object in the customer to be recommended; the distance function is used to define the proximity between each cluster object and each first cluster center value;
[0087] Calculate the mean of each updated cluster group, and use each mean as the second cluster center value of each updated cluster group;
[0088] The second cluster center value is used as the cluster center value of the next iteration, and the iterative clustering is repeated until the cluster center value of each cluster group reaches stability, then the iteration is stopped to obtain a plurality of preset numbers of customer groups.
[0089] First, it is necessary to determine the number of clusters, that is, to determine how many customer groups the customer group to be recommended should be divided into; the number of clusters is determined according to the aforementioned specific business needs and goals and the dimensions of the customer data to be recommended. Then select the first cluster center value of each cluster group; as an example, for each cluster group, a point can be selected as the initial cluster center value by random selection; as another example, a point can be randomly selected as the initial cluster center of the first cluster group, and then the point farthest from the point is selected as the initial cluster center of the second cluster group; the point farthest from the selected center point of the previous cluster group is selected as the initial cluster center of the next cluster group in turn, until the initial cluster centers of all cluster groups are selected. Next, multiple updated cluster groups are obtained according to the distance function; wherein the distance function represents the degree of proximity between each cluster object and each first cluster center value. After obtaining multiple updated cluster groups, the mean of the cluster objects in each cluster group is calculated, and the mean corresponding to each cluster group is used as the center value of the next clustering; the iteration is repeated until the cluster center value of each cluster group reaches stability, and then the iteration is stopped; as an example, when the iteration reaches the preset number of iterations, the iteration is stopped. After the iteration is completed, multiple preset numbers of customer groups can be obtained. In this example, the clustering algorithm is used to cluster the recommended customer groups, which improves the accuracy of customer group classification.
[0090] Further, according to each first cluster center value and the cluster objects in each cluster group, each updated cluster group is obtained based on a distance function, including:
[0091] Calculate the distance between each cluster object and each first cluster center value, and select the first cluster center value with the shortest distance to each cluster object as the first cluster center value of the cluster group where the cluster object belongs;
[0092] The clustering object under the cluster group where the first cluster center value is located is determined according to each first cluster center value to obtain an updated cluster group.
[0093] According to the first cluster center value of each cluster group and each cluster object, the distance between each cluster object (that is, each data object) in the entire customer group to be recommended and the initial first cluster center value corresponding to each cluster group is calculated; the first cluster center with the smallest distance is selected as the center value of the data object; according to each first cluster center value, the cluster object under the cluster group where the first cluster center value is located is determined to obtain the updated cluster group. In this example, by calculating the distance between each cluster object and each center value, the center value corresponding to the cluster object with the smallest distance to it is selected as the new center value, which further improves the accuracy of the classification of the customer group to be recommended.
[0094] After the customer groups to be recommended are clustered to obtain multiple cluster sets, it is necessary to recommend credit card products to the customer groups to be recommended; the recommendation module 22 is used to:
[0095] Calculating the fitness of each product to be recommended and each second customer group; the fitness represents the matching degree of each product to be recommended and each second customer group;
[0096] According to the fitness, target recommended products are selected based on the products to be recommended.
[0097] Specifically, first set a fitness evaluation criterion, which can be formulated according to actual business needs; calculate the fitness of each product to be recommended and each second customer group; where the fitness represents the degree of match between each product to be recommended and the second customer group; the higher the fitness value, the higher the degree of match between the product to be recommended and the evaluation criteria; select the target recommended product from the products to be recommended according to the fitness. In one example, a particle swarm algorithm is used to select the target recommended product; in actual applications, the particle swarm algorithm is an intelligent optimization search method based on the group; the particle swarm optimization algorithm searches the solution space by continuously updating the speed and position of the particles, and each particle updates its speed and position based on its own experience and the experience of the group; specifically, each particle calculates the new speed and position based on the current speed and position and its own historical optimal position and the historical optimal position of the group. Through continuous iterative updates, the particle swarm gradually approaches the global optimal solution. The usual steps of using the particle swarm optimization algorithm are: first initialize the population, then calculate the fitness of each object in the population (the product to be recommended in this application); update the particle speed and particle position according to the fitness; calculate the fitness at the new position, if the fitness of the new position is higher, then update the position of the particle, otherwise do not update; repeat the iteration until the termination condition is met; then output the optimal value, which is the target recommended product of this application. This example uses the optimization algorithm to select the target recommended product according to the fitness, so that the recommendation results of credit card products are more accurate.
[0098] Furthermore, according to the fitness, target recommended products are selected based on the products to be recommended, including:
[0099] According to the fitness, the recommended products with fitness higher than the preset value are selected as the target recommended products.
[0100] In actual applications, based on the fitness of each to-be-recommended product and each second customer group, the to-be-recommended product with a fitness higher than a preset value is selected as the target recommended product; the target recommended product is subsequently recommended to the corresponding second customer group. As an example, there may be multiple target recommended products, and multiple target recommended products are recommended to the user, and the customer can subsequently select products from the multiple target recommended products. As another example, multiple target recommended products can be sorted according to fitness, and credit card products can be recommended to the recommended customers according to priority based on the sorting results. In this example, the target recommended product corresponding to the second customer group is selected according to the optimization algorithm and recommended to the second customer group, thereby improving the accuracy of credit card product recommendations.
[0101] The credit card product recommendation device provided in this embodiment obtains a first customer group by clustering the credit card behavior data of the customer group to be recommended using a clustering algorithm; then, based on the obtained personal basic information data and financial data of the first customer group, a clustering algorithm is used again to cluster to obtain multiple second customer groups; then, based on the corresponding products to be recommended of each second customer group, an optimization algorithm is used to obtain the target recommended products corresponding to each second customer group. The solution of this application accurately classifies the customer group to be recommended by clustering the customer group to be recommended twice, and then recommends products to the second customer group using an optimization algorithm based on multiple products to be recommended; the solution of this application improves the accuracy of credit card product recommendations.
[0102] Embodiment 3
[0103] Figure 6 exemplarily shows a structural schematic diagram of an electronic device, the device comprising:
[0104] The device includes a processor 291 and a memory 292; it may also include a communication interface 293 and a bus 294. The processor 291, the memory 292, and the communication interface 293 may communicate with each other through the bus 294. The communication interface 293 may be used for information transmission. The processor 291 may call the logic instructions in the memory 292 to execute the above-mentioned example method.
[0105] In addition, the logic instructions in the above-mentioned memory 292 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.
[0106] The memory 292 is a computer-readable storage medium that can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the methods in the embodiments of the present application. The processor 291 executes functional applications and data processing by running the software programs, instructions, and modules stored in the memory 292, that is, implementing the methods in the above method examples.
[0107] The memory 292 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 292 may include a high-speed random access memory and may also include a non-volatile memory.
[0108] An embodiment of the present application further provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method in any embodiment.
[0109] An embodiment of the present application also provides a computer program product, including a computer program, which is used to implement the method in any embodiment when executed by a processor.
[0110] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.
[0111] It should be further noted that, although the various steps in the flowchart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
[0112] It should be understood that the above-mentioned device embodiments are only illustrative, and the device of the present application can also be implemented in other ways. For example, the division of units / modules in the above-mentioned embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.
[0113] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present application may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.
[0114] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. Unless otherwise specified, the processor may be any appropriate hardware processor, such as a CPU, a GPU, an FPGA, a DSP, an ASIC, etc. Unless otherwise specified, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory (RRAM), a dynamic random access memory (DRAM), a static random access memory (SRAM), an enhanced dynamic random access memory (EDRAM), a high-bandwidth memory (HBM), a hybrid memory cube (HMC), etc.
[0115] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including a number of 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 various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.
[0116] In the above embodiments, the description of each embodiment has its own emphasis. For the part not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0117] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.
[0118] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.
Claims
1. A credit card product recommendation method, characterized in that: The method comprises: According to the first historical information of the customer group to be recommended, clustering the customer group to be recommended using a clustering algorithm to obtain a first customer group; the first historical information includes credit card user behavior data of the customer group to be recommended; According to the second historical information of the first customer group, clustering the first customer group using a clustering algorithm to obtain a plurality of second customer groups; the second historical information includes basic personal information data and financial information data of the first customer group; According to the to-be-recommended products corresponding to each second customer group, an optimization algorithm is used to obtain target recommended products for each second customer group; and the target recommended products are sent to the corresponding second customer group.
2. The method according to claim 1, characterized in that The step of clustering the customer group to be recommended using a clustering algorithm to obtain a first customer group according to the first historical information of the customer group to be recommended includes: According to the first historical information, the customer group to be recommended is classified by using the clustering algorithm to obtain a plurality of customer groups of a preset number; A customer group whose credit card behavior data is higher than a preset value is selected from multiple customer groups as the first customer group.
3. The method according to claim 2, characterized in that The step of classifying the customer group to be recommended by using the clustering algorithm to obtain a plurality of customer groups of a preset number according to the first historical information includes: Acquire the first historical information; According to the target attributes, feature extraction is performed on the first historical information to obtain feature data; the target attributes include the number of credit card participations, the number of shares, and the number of conversions; According to the characteristic data, clustering processing is performed using the clustering algorithm to obtain the plurality of preset numbers of customer groups.
4. The method according to any one of claims 1 to 3, characterized in that: The step of using a clustering algorithm to perform clustering processing to obtain the plurality of preset number of customer groups includes: Determine the number of clusters; Randomly select the first cluster center value of each cluster group; According to the first cluster center values and the cluster objects in each cluster group, each updated cluster group is obtained based on a distance function; the cluster objects represent each customer object in the to-be-recommended customers; the distance function is used to define the proximity between each cluster object and each first cluster center value; Calculate the mean of each updated cluster group, and use each mean as the second cluster center value of each updated cluster group; The second cluster center value is used as the cluster center value of the next iteration, and the iterative clustering is repeated until the cluster center value of each cluster group reaches stability, then the iteration is stopped to obtain the plurality of preset numbers of customer groups.
5. The method according to claim 4, characterized in that The step of obtaining each updated cluster group based on a distance function according to each first cluster center value and the cluster objects in each cluster group includes: Calculate the distance between each cluster object and each first cluster center value, and select the first cluster center value with the shortest distance to each cluster object as the first cluster center value of the cluster group where the cluster object belongs; The cluster objects under the cluster group where the first cluster center value is located are determined according to each first cluster center value to obtain the updated cluster group.
6. The method according to any one of claims 1 to 3, characterized in that: The method of using the optimization algorithm to obtain the target recommended product for each second customer group includes: Calculating the fitness of each of the products to be recommended and each of the second customer groups; the fitness represents the degree of matching between each of the products to be recommended and each of the second customer groups; According to the fitness, the target recommended product is selected based on the product to be recommended.
7. The method according to claim 6, characterized in that The selecting the target recommended product based on the product to be recommended according to the fitness includes: According to the fitness, the to-be-recommended product whose fitness is higher than a preset value is selected as the target recommended product.
8. A credit card product recommendation device, characterized in that: The device comprises: A classification module, configured to cluster the customer group to be recommended using a clustering algorithm to obtain a first customer group according to first historical information of the customer group to be recommended; the first historical information includes credit card user behavior data of the customer group to be recommended; The classification module is further used to cluster the first customer group using a clustering algorithm to obtain a plurality of second customer groups according to the second historical information of the first customer group; the second historical information includes basic personal information data and financial information data of the first customer group; The recommendation module is used to obtain target recommended products for each second customer group by using an optimization algorithm according to the products to be recommended corresponding to each second customer group; and send the target recommended products to the corresponding second customer group.
9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.
11. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.