Adjusting method and adjusting device of collection strategy and electronic equipment

Abstract

The invention provides a collection strategy adjusting method and device and electronic equipment. The method comprises the following steps: acquiring historical data of a plurality of customers; dividing the plurality of clients into different client groups by using a clustering algorithm according to the historical data; at least utilizing the decision tree model to generate a plurality of collection strategies for each customer group; executing a plurality of collection strategies on each customer group to obtain a plurality of collection results, and evaluating the effects of the plurality of collection results by using an adaptive algorithm; adjusting the proportion of a plurality of collection strategies in the customer group at least according to the effect to obtain a plurality of adjusted execution proportions; dividing clients in the client group to obtain a plurality of client subgroups; and adjusting the collection strategy of each customer subgroup into the corresponding collection strategy. The method solves the problem that the collection strategy cannot be adaptively adjusted in the prior art.

Description

Technical Field

[0001] This application relates to the field of debt collection technology. Specifically, it relates to a method and device for adjusting debt collection strategies, a computer-readable storage medium, a processor, and an electronic device. Background Art

[0002] In the financial field, debt collection is a crucial task. Traditional debt collection methods usually adopt fixed strategies, lack personalized adjustment, and are difficult to cope with changes in different customers and market environments. In addition, traditional evaluation methods may not be able to comprehensively and accurately measure the effectiveness of debt collection strategies, resulting in instability in debt collection effects.

[0003] Traditional implementation solutions usually adopt fixed debt collection strategies, which remain unchanged throughout the debt collection cycle. These strategies are usually based on manual experience and rules, and may make some simple adjustments according to the customer's historical behavior and default situation. The core idea of these traditional methods is to decide what method to adopt for debt collection based on the customer's default situation: Rule-based strategies: Traditional methods usually use rule-based debt collection strategies. For example, for customers with a higher default risk, the frequency of phone calls for debt collection can be increased, the number of debt collection letters sent, etc. These rules are usually static and rarely adjusted once formulated, and cannot cope with changes in the market and customer behavior; Phone, letter, and SMS notifications: Traditional methods usually notify customers to make repayments through phone calls, letters, SMS, etc. For different customers, the appropriate notification method can be selected according to their contact information and preferences; Manual decision-making and experience: In traditional methods, debt collection personnel usually formulate debt collection strategies based on experience and intuition, which may lead to differences in strategies among different debt collection personnel and lack unified and quantitative standards.

[0004] Although traditional implementation solutions may be effective in some cases, they have limitations such as static strategies, lack of personalization, and rule-based, and are difficult to cope with the challenges of dynamic changes.

[0005] Therefore, there is an urgent need for a method that can adaptively adjust debt collection strategies. Summary of the Invention

[0006] The main purpose of this application is to provide a method and device for adjusting debt collection strategies, a computer-readable storage medium, a processor, and an electronic device, so as to at least solve the problem that the prior art cannot adaptively adjust debt collection strategies.

[0007] To achieve the above object, according to one aspect of the present application, a method for adjusting a collection strategy is provided, including: an acquisition step of acquiring historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records; a first partitioning step of partitioning the multiple customers into different customer groups according to the historical data by using a clustering algorithm; a generation step of generating multiple collection strategies for each customer group by at least using a decision tree model, where the collection strategy is to use different collection methods every day within a preset time, and the collection methods at least include making phone calls and sending text messages; an evaluation step of executing the multiple collection strategies for each customer group to obtain multiple collection results, and using an adaptive algorithm to evaluate the effects of the multiple collection results; a first adjustment step of adjusting the proportions of the multiple collection strategies in the customer group at least according to the effects to obtain multiple adjusted execution proportions; a second partitioning step of partitioning the customers in the customer group to obtain multiple customer subgroups, and the proportions between the multiple customer subgroups are equal to the execution proportions; a second adjustment step of adjusting the collection strategies of each customer subgroup to the corresponding collection strategies, the ratio of each customer subgroup to the customer group is a predetermined proportion, and the ratio of the proportion of the collection strategies corresponding to each customer subgroup to the sum of the proportions of all the collection strategies is the predetermined proportion.

[0008] Optionally, the generation step includes: inputting the historical data of the multiple customers in each customer group into the decision tree model to obtain multiple initial collection strategies; using a random forest algorithm to at least screen the multiple initial collection strategies to obtain the multiple collection strategies.

[0009] Optionally, the first adjustment step includes: scoring the multiple collection strategies according to the effects of the multiple collection strategies to obtain multiple scores, where the effects include the results of multiple evaluation indicators, and the evaluation indicators at least include the recovery rate and the default rate; determining the optimal collection strategy and the worst collection strategy according to the scores; at least increasing the first proportion of the optimal collection strategy and decreasing the second proportion of the worst collection strategy to obtain the multiple adjusted execution proportions.

[0010] Optionally, scoring the multiple collection strategies according to the effects of the multiple collection strategies to obtain multiple scores includes: normalizing the effects of each collection strategy respectively to obtain multiple evaluation index data; according to y = a 1 x 1 +a 2 x 2 +......+a n x n , scoring each collection strategy to obtain the multiple scores y, where x1 , x 2 ,......x n For multiple evaluation index data, a 1 , a 2 ,......a n are the weights corresponding to the multiple evaluation index data.

[0011] Optionally, according to the scores, determine the optimal collection strategy and the worst collection strategy, including: determining the collection strategy corresponding to the maximum score as the optimal collection strategy; determining the collection strategy corresponding to the minimum score as the worst collection strategy.

[0012] Optionally, before at least adjusting the first ratio of the optimal collection strategy and the second ratio of the worst collection strategy to obtain the adjusted multiple execution ratios, the method further includes: determining whether the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy is within a preset range; when the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy is within the preset range, at least adjust the parameters of the clustering algorithm, and repeat the first partitioning step, the generating step, and the evaluating step at least once until the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy is not within the preset range.

[0013] Optionally, at least adjusting the first ratio of the optimal collection strategy and the second ratio of the worst collection strategy to obtain the adjusted multiple execution ratios, including: increasing the first ratio of the optimal collection strategy by a first preset ratio to obtain an adjusted first execution ratio; reducing the second ratio of the worst collection strategy by a second preset ratio to obtain an adjusted second execution ratio, where the first preset ratio is equal to the second preset ratio, and the first ratio is equal to the second ratio.

[0014] According to another aspect of the present application, there is provided an adjustment device for a collection strategy, including:

[0015] An acquisition unit, configured to execute the acquisition step to acquire historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction flow;

[0016] A first partitioning unit, configured to execute the first partitioning step to partition multiple customers into different customer groups according to the historical data by using a clustering algorithm;

[0017] A generation unit, configured to perform a generation step, and generate, for each of the customer groups, a plurality of collection strategies by using at least a decision tree model, where the collection strategies are different collection methods used every day within a preset time, and the collection methods at least include making calls and sending text messages;

[0018] An evaluation unit, configured to perform an evaluation step, execute the plurality of collection strategies for each of the customer groups, obtain a plurality of collection results, and evaluate the effects of the plurality of collection results by using an adaptive algorithm;

[0019] A first adjustment unit, configured to perform a first adjustment step, and adjust, at least according to the effects, the proportions of the plurality of collection strategies in the customer groups to obtain adjusted plurality of execution proportions;

[0020] A second division unit, configured to perform a second division step, divide the customers in the customer groups to obtain a plurality of customer subgroups, and the proportions between the plurality of customer subgroups are equal to the execution proportions;

[0021] A second adjustment unit, configured to perform a second adjustment step, adjust the collection strategies of each of the customer subgroups to the corresponding collection strategies, the ratio of each customer subgroup to the customer group is a predetermined proportion, and the ratio of the proportion of the collection strategies corresponding to each customer subgroup to the sum of the proportions of all the collection strategies is the predetermined proportion.

[0022] According to another aspect of the present application, there is provided a computer-readable storage medium, where the computer-readable storage medium includes a stored program, and when the program runs, it controls a device where the computer-readable storage medium is located to execute any one of the above-mentioned collection strategy adjustment methods.

[0023] According to still another aspect of the present application, there is provided a processor, where the processor is used to run a program, and when the program runs, it executes any one of the above-mentioned collection strategy adjustment methods.

[0024] According to still another aspect of the present application, there is provided an electronic device, including: one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include those for executing any one of the above-mentioned collection strategy adjustment methods.

[0025] Applying the technical solution of the present application, first, obtain the historical data of multiple customers; secondly, according to the historical data of the customers, use the clustering algorithm to divide the multiple customers into different customer groups; then, generate multiple collection strategies for each customer group at least using the decision tree model; execute the multiple collection strategies, and use the adaptive algorithm to evaluate the effect of the generated results after executing the collection strategies; according to the effect, adjust the proportion of the collection strategies executed in the customer group, and divide the customers in the customer group according to the proportion to obtain customer subgroups; finally, adjust the collection strategies of each customer subgroup to the collection strategies with the same proportion. Compared with the prior art, the solution of the present application uses the adaptive algorithm to monitor in real time the collection execution results of customer groups with different characteristics divided by the clustering algorithm. According to the collection execution results, dynamically adjust the proportion of the collection strategies. According to the proportion situation, the collection strategies of the customer group will be gradually adjusted to the collection strategy with the largest proportion, that is, this method uses the clustering algorithm to subdivide customers into different customer groups according to customer characteristics, and according to the results of different collection strategies generated by the decision tree for the customer group, at least use the adaptive algorithm to dynamically adjust the execution proportion of the collection strategies of each customer group, forming a dynamic feedback process. According to the continuously changing customer characteristics, adjust the execution proportion of the collection strategies of the customer group divided by customer characteristics, realizing the adaptive adjustment of the collection strategies, thus solving the problem that the collection strategies cannot be adaptively adjusted in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings forming a part of this application are used to provide a further understanding of the application. The schematic embodiments and descriptions thereof of the application are used to explain the application and do not constitute an improper limitation of the application. In the drawings:

[0027] Figure 1 FIG. shows a hardware structure block diagram of a mobile terminal for an adjustment method of executing a collection strategy provided in an embodiment of the present application;

[0028] Figure 2 FIG. shows a schematic flow chart of an adjustment method of a collection strategy provided in an embodiment of the present application;

[0029] Figure 3 FIG. shows a structure block diagram of an adjustment device of a collection strategy provided in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0031] To enable those skilled in the art to better understand the solution of this application, the following will clearly and completely describe the technical solution in the embodiments of this application in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned accompanying drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances for the embodiments of this application described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0033] As introduced in the background art, there is a problem in the prior art that the collection strategy cannot be adaptively adjusted. To solve the above problem, the embodiments of this application provide a method for adjusting a collection strategy, an adjustment device, a computer-readable storage medium, a processor, and an electronic device.

[0034] The following will clearly and completely describe the technical solution in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention.

[0035] The method embodiments provided in the embodiments of this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 is a block diagram of the hardware structure of a mobile terminal for a method of adjusting a collection strategy according to an embodiment of the present invention. As Figure 1 shown, the mobile terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Among them, the above-mentioned mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned mobile terminal. For example, the mobile terminal may further include more or fewer components than those shown in Figure 1 the figure, or have a different configuration from that shown in Figure 1 the figure.

[0036] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiments of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implements the above method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories can be connected to the mobile terminal through a network. Examples of the above network include but are not limited to the Internet, intranet, local area network, mobile communication network, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0037] In this embodiment, a method for adjusting a collection strategy running on a mobile terminal, a computer terminal, or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0038] Figure 2 is a flowchart of the method for adjusting the collection strategy according to the embodiments of the present application. As Figure 2 shown, the method includes the following steps:

[0039] Step S201, an acquisition step, to acquire historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records.

[0040] Specifically, collect customer information from the financial system, clean the acquired data, process missing values and outliers to ensure the accuracy and integrity of the data, and integrate and standardize the data so that information from different data sources can be used uniformly. Among them, customer information includes: basic information such as name, ID number, contact information, etc.; income situation includes data covering aspects such as salary and other income; transaction records include bank account, credit card transactions, etc.

[0041] Step S202, the first partitioning step, according to the above historical data, using a clustering algorithm to partition the multiple above-mentioned customers into different customer groups.

[0042] Specifically, use a clustering algorithm to analyze customer data, partition customers into multiple groups according to different characteristics. This algorithm needs to determine an appropriate number of clusters to achieve the subdivision of the customer dimension, and for each group, analyze its common characteristics and behavior patterns in order to generate appropriate collection strategies for them subsequently.

[0043] Step S203, the generation step, at least use a decision tree model to generate multiple collection strategies for each of the above-mentioned customer groups. The above-mentioned collection strategies are to use different collection methods every day within a preset time. The above-mentioned collection methods at least include making phone calls and sending text messages.

[0044] Specifically, based on the characteristics of each customer group and historical collection data, at least use a decision tree to generate multiple collection strategies; generate multiple strategies for each group, including collection letters, phone notifications, text message reminders, etc., to adapt to the preferences and communication methods of different customers.

[0045] Step S204, the evaluation step, execute multiple above-mentioned collection strategies for each of the above-mentioned customer groups to obtain multiple collection results, and use an adaptive algorithm to evaluate the effects of the multiple above-mentioned collection results.

[0046] Specifically, execute collection actions according to the generated collection strategies, including sending collection letters, making phone calls, etc., and obtain the time, method, result of each collection action, as well as the response of the customer; collect repayment information, extension situations, etc., to supplement the result data of collection execution.

[0047] Step S205, the first adjustment step, at least adjust the proportions of multiple above-mentioned collection strategies in the above-mentioned customer groups according to the above-mentioned effects to obtain multiple adjusted execution proportions.

[0048] Specifically, according to the evaluated effects, adjust the proportion of different collection strategies within each customer group to optimize the overall collection effect.

[0049] Step S206, the second partitioning step, partition the customers in the above-mentioned customer groups to obtain multiple customer subgroups, and the proportion between the multiple above-mentioned customer subgroups is equal to the above-mentioned execution proportion.

[0050] Step S207, the second adjustment step, adjust the above-mentioned collection strategies of each of the above-mentioned customer subgroups to the corresponding above-mentioned collection strategies. The ratio of each of the above-mentioned customer subgroups to the above-mentioned customer group is a predetermined proportion, and the ratio of the proportion of the above-mentioned collection strategies corresponding to each of the above-mentioned customer subgroups to the sum of the proportions of all the above-mentioned collection strategies is the above-mentioned predetermined proportion.

[0051] In the above embodiments, first, historical data of multiple customers is obtained; second, according to the historical data of the customers, multiple customers are divided into different customer groups by using a clustering algorithm; then, at least a decision tree model is used to generate multiple collection strategies for each customer group; the multiple collection strategies are executed, and an adaptive algorithm is used to evaluate the effect of the results generated after executing the collection strategies; according to the effect, the proportion of executing the collection strategies in the customer group is adjusted, and the customers in the customer group are divided according to the proportion to obtain customer subgroups; finally, the collection strategies of each customer subgroup are adjusted to the collection strategies with the same proportion. Compared with the prior art, the solution of the present application uses an adaptive algorithm to monitor in real time the collection execution results of customer groups with different characteristics divided by the clustering algorithm. According to the collection execution results, the proportion of the collection strategies is dynamically adjusted. According to the proportion situation, the collection strategies of the customer group will be gradually adjusted to the collection strategy with the largest proportion, that is, this method uses the clustering algorithm to subdivide customers into different customer groups according to customer characteristics, and according to the results of different collection strategies generated by the decision tree for the customer groups, at least the adaptive algorithm is used to dynamically adjust the execution proportion of the collection strategies of each customer group, forming a dynamic feedback process. According to the continuously changing customer characteristics, the execution proportion of the collection strategies of the customer groups divided by customer characteristics is adjusted, realizing the adaptive adjustment of the collection strategies, thus solving the problem that the collection strategies cannot be adaptively adjusted in the prior art.

[0052] In an embodiment of the present application, the generating step includes: inputting the historical data of multiple customers in each of the above customer groups into the above decision tree model to obtain multiple initial collection strategies; using a random forest algorithm to at least screen the multiple initial collection strategies to obtain the multiple collection strategies. The random forest algorithm in this method can combine the results of multiple decision tree models, reduce the risk of overfitting, further improve the accuracy and stability of the model, and thus make the formulation of the collection strategies more accurate.

[0053] Specifically, according to the decision tree model, by inputting the characteristics of a customer group and following the path from the root node to the leaf node, multiple initial collection strategies for the group can be obtained. The random forest model can generate multiple collection strategies. Each strategy is based on the output of a single decision tree. Through majority voting or weighted averaging, multiple collection strategies can be generated for each customer group.

[0054] In order to more accurately adjust the execution ratios of multiple collection strategies in a customer group, and thus more accurately adjust the collection strategies, in an embodiment of the present application, the first adjustment step includes: scoring the multiple collection strategies according to the effects of the multiple collection strategies to obtain multiple scores, where the effects include the results of multiple evaluation indicators, and the evaluation indicators at least include the recovery rate and the default rate; determining the optimal collection strategy and the worst collection strategy according to the scores; and increasing at least a first ratio of the optimal collection strategy and decreasing a second ratio of the worst collection strategy to obtain the adjusted multiple execution ratios.

[0055] In another embodiment, scoring the multiple collection strategies according to the effects of the multiple collection strategies to obtain multiple scores includes: normalizing the effects of each collection strategy respectively to obtain multiple evaluation index data; and scoring each collection strategy according to y = a 1 x 1 +a 2 x 2 +......+a n x n to obtain multiple scores y, where x 1 , x 2 ,......x n are multiple evaluation index data, and a 1 , a 2 ,......a n are the weights corresponding to the multiple evaluation index data. The magnitudes of the weights are determined according to the importance of the evaluation indicators, that is, determined according to the influence degree of the evaluation indicators on the collection strategies. The importance is directly proportional to the magnitudes of the weights. In this method, by normalizing the effects of each collection strategy, the dimensional difference between different evaluation indicators can be further eliminated, enabling them to be compared. Then, the weighted sum of each evaluation index data is calculated according to the weights to obtain the final score, so that the score can be made more accurate, and further the collection strategy can be more accurately and adaptively adjusted.

[0056] Specifically, the purpose of normalization is mainly to solve the comparison problems caused by different dimensions and ranges of evaluation indicators. For example, the collection rate may be a percentage, while the collection cost may be measured in monetary units. It is unfair to directly compare the original values of these indicators. Normalization can convert all indicators to a common scale, enabling the data corresponding to the effects to be directly used for comparison between strategies. There are two common normalization methods: Min-Max Scaling scales all data to the range of 0 to 1; Z-score Standardization converts the data into a standard normal distribution, that is, the mean is 0 and the standard deviation is 1.

[0057] In order to more accurately determine the optimal collection strategy and the worst collection strategy, so that the execution ratio of the adjusted collection strategy is more accurate. In one embodiment of the present application, according to the above scores, the optimal collection strategy and the worst collection strategy are determined, including: determining the collection strategy corresponding to the maximum above score as the above optimal collection strategy; determining the collection strategy corresponding to the minimum above score as the above worst collection strategy.

[0058] In another embodiment, before at least adjusting the first ratio of the above optimal collection strategy and the second ratio of the above worst collection strategy to obtain the adjusted multiple above execution ratios, the method further includes: judging whether the absolute value of the difference between the first score of the above optimal collection strategy and the second score of the above worst collection strategy is within a preset range; when the absolute value of the difference between the first score of the above optimal collection strategy and the second score of the above worst collection strategy is within the above preset range, at least adjust the parameters of the above clustering algorithm, and repeat the above first partitioning step, the above generating step, and the above evaluating step at least once until the absolute value of the difference between the first score of the above optimal collection strategy and the second score of the above worst collection strategy is not within the above preset range. In this method, by continuously adjusting the execution ratios of the optimal collection strategy and the worst collection strategy and adjusting the parameters of the clustering algorithm, a better combination of collection strategies is further searched. By judging whether the absolute value of the score difference between the optimal collection strategy and the worst collection strategy is within the preset range, it can be further determined whether the current strategy combination is effective, so that the execution ratio of the collection strategy is more accurate, and thus the collection strategy can be adjusted more accurately.

[0059] Specifically, first, the system calculates the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy. This absolute value of the difference reflects the maximum difference in the strategy effect. Set a preset range, such as a score difference of 0.1 to 0.5. If the difference between the optimal strategy and the worst strategy is within this range, it is considered that the current strategy has limited room for optimization, and more detailed customer group segmentation or strategy generation may be required; second, if the score difference between the optimal and worst strategies is within the preset range, the system will re-examine the accuracy of the customer group segmentation and optimize the group segmentation by adjusting the parameters of the clustering algorithm. For example, the number of clusters, the initialization method of the cluster centers, the distance metric standard, etc. can be adjusted to obtain a more accurate customer group that can better distinguish the strategy effects; after that, after adjusting the parameters of the clustering algorithm, the system will re-execute the first segmentation step of the customer group segmentation, and then generate a new collection strategy using the decision tree model according to the newly segmented groups, that is, the generation step. Then, execute the strategy evaluation step, collect the strategy execution results, including indicators such as the repayment rate and the default rate, and re-rate each strategy; finally, the system determines again whether the score difference between the optimal and worst strategies is within the preset range. If not, it means that the strategy effect difference is significant and the strategy ratio adjustment stage can be entered. If it is still within the preset range, the system will continue to adjust the parameters of the clustering algorithm and repeat the above segmentation, generation, and evaluation steps until the score difference between the optimal and worst strategies exceeds the preset range, that is, the strategy effect difference is significant enough to perform strategy adjustment.

[0060] In order to keep the sum of the execution ratios of the optimal collection strategy and the worst collection strategy unchanged, so as to make the distribution of the collection strategies more balanced, in one embodiment of the present application, at least adjust the first ratio of the above optimal collection strategy and the second ratio of the above worst collection strategy to obtain multiple adjusted above execution ratios, including: increasing the above first ratio of the above optimal collection strategy by a first preset ratio to obtain an adjusted first execution ratio; reducing the above second ratio of the above worst collection strategy by a second preset ratio to obtain an adjusted second execution ratio, the above first preset ratio is equal to the above second preset ratio, and the above first ratio is equal to the above second ratio.

[0061] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the adjustment method of the collection strategy of the present application will be described in detail below in conjunction with specific embodiments.

[0062] This embodiment relates to a specific adjustment method of a collection strategy, including the following steps:

[0063] Step 1: Obtain the historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records;

[0064] Step 2: According to the historical data of customers, use the clustering algorithm to divide customers into different customer groups;

[0065] Step 3: Use the decision tree to generate multiple initial collection strategies for each customer group, and use the random forest algorithm to screen multiple initial collection strategies to obtain multiple collection strategies;

[0066] Step 4: Execute multiple collection strategies on the customer groups to obtain multiple collection results, and use the adaptive algorithm to evaluate the effects of multiple collection results to obtain multiple evaluation index data, where the effects include at least the recovery rate and the default rate;

[0067] Step 5: According to the effects, use y = a 1 x 1 + a 2 x 2 +......+ a n x n , score multiple collection strategies to obtain multiple scores, where x 1 , x 2 ,......x n are multiple evaluation index data, and a 1 , a 2 ,......a n are the weights corresponding to multiple above-mentioned evaluation index data. Determine the one with the highest score as the optimal collection strategy and the one with the lowest score as the worst collection strategy. Judge whether the difference between the score of the optimal collection strategy and the score of the worst collection strategy is within the preset range. If it is within the preset range, at least adjust the parameters in the clustering algorithm, and repeat Step 2, Step 3, and Step 4 at least once until the difference between the two is not within the preset range. Increase the proportion of the optimal collection strategy by a preset proportion and decrease the proportion of the worst collection strategy by a preset proportion to obtain multiple adjusted execution proportions;

[0068] Step 6: Divide the customers in the customer group according to the execution proportion to obtain multiple customer subgroups;

[0069] Step 7: According to the predetermined ratio, adjust the collection strategy of the customer subgroup to the corresponding collection strategy, where the ratio of the customer subgroup to the customer group is the predetermined ratio, and the ratio of the proportion of the collection strategy corresponding to the customer subgroup to the sum of the proportions of all collection strategies is the predetermined ratio.

[0070] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0071] The embodiments of the present application also provide an adjustment device for collection strategies. It should be noted that the adjustment device for collection strategies in the embodiments of the present application can be used to execute the method for adjusting collection strategies provided in the embodiments of the present application. The device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0072] The following introduces the adjustment device for collection strategies provided in the embodiments of the present application.

[0073] Figure 3 is a schematic diagram of the adjustment device for collection strategies according to the embodiments of the present application. As Figure 3 shown, the device includes:

[0074] An acquisition unit 10, configured to execute an acquisition step to acquire historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records.

[0075] Specifically, collect customer information from the financial system, clean the acquired data, process missing values and outliers to ensure the accuracy and integrity of the data, and integrate and standardize the data so that information from different data sources can be used uniformly. Among them, customer information includes: basic information such as name, ID number, contact information, etc.; income situation is data covering aspects such as salary and other income; transaction records are bank account, credit card transactions, etc.

[0076] A first division unit 20, configured to execute a first division step to divide multiple customers into different customer groups according to the historical data by using a clustering algorithm.

[0077] Specifically, use a clustering algorithm to analyze customer data, divide customers into multiple groups according to different characteristics. This algorithm needs to determine an appropriate number of clusters to achieve the subdivision of the customer dimension, and for each group, analyze its common characteristics and behavior patterns in order to generate appropriate collection strategies for it subsequently.

[0078] A generation unit 30, configured to execute a generation step to generate multiple collection strategies for each customer group at least by using a decision tree model. The collection strategies are to use different collection methods every day within a preset time, and the collection methods at least include making calls and sending text messages.

[0079] Specifically, based on the characteristics of each customer group and historical collection data, multiple collection strategies are generated at least using decision trees; multiple strategies are generated for each group, including collection letters, phone notifications, text message reminders, etc., to adapt to the preferences and communication methods of different customers.

[0080] The evaluation unit 40 is used to perform the evaluation step, execute multiple above-mentioned collection strategies for each of the above-mentioned customer groups, obtain multiple collection results, and use an adaptive algorithm to evaluate the effects of the multiple above-mentioned collection results.

[0081] Specifically, collection actions are carried out in accordance with the generated collection strategies, including sending collection letters, making phone calls, etc., and the time, method, result of each collection action, and the customer's response are obtained; repayment information, extension situations, etc. are collected to supplement the result data of collection execution.

[0082] The first adjustment unit 50 is used to perform the first adjustment step, adjust the proportions of multiple above-mentioned collection strategies in the above-mentioned customer groups at least according to the above-mentioned effects, and obtain multiple adjusted execution proportions.

[0083] Specifically, according to the evaluated effects, the proportions of different collection strategies within each customer group are adjusted to optimize the overall collection effect.

[0084] The second division unit 60 is used to perform the second division step, divide the customers in the above-mentioned customer groups to obtain multiple customer subgroups, and the proportion between the multiple above-mentioned customer subgroups is equal to the above-mentioned execution proportion.

[0085] The second adjustment unit 70 is used to perform the second adjustment step, adjust the above-mentioned collection strategies of each of the above-mentioned customer subgroups to the corresponding above-mentioned collection strategies, the ratio of each of the above-mentioned customer subgroups to the above-mentioned customer group is a predetermined proportion, and the ratio of the proportion of the above-mentioned collection strategies corresponding to each of the above-mentioned customer subgroups to the sum of the proportions of all the above-mentioned collection strategies is the above-mentioned predetermined proportion.

[0086] In the above embodiments, first, historical data of multiple customers is obtained; second, according to the historical data of the customers, multiple customers are divided into different customer groups by using a clustering algorithm; then, at least a decision tree model is used to generate multiple collection strategies for each customer group; the multiple collection strategies are executed, and an adaptive algorithm is used to evaluate the effect of the generated results after executing the collection strategies; according to the effect, the proportion of executing the collection strategies in the customer group is adjusted, and the customers in the customer group are divided according to the proportion to obtain customer subgroups; finally, the collection strategies of each customer subgroup are adjusted to the collection strategies with the same proportion as theirs. Compared with the prior art, the solution of the present application uses an adaptive algorithm to monitor in real time the collection execution results of customer groups with different characteristics divided by the clustering algorithm. According to the collection execution results, the proportion of the collection strategies is dynamically adjusted. According to the proportion situation, the collection strategies of the customer group will be gradually adjusted to the collection strategy with the largest proportion, that is, this method uses the clustering algorithm to subdivide customers into different customer groups according to customer characteristics, and according to the results of different collection strategies generated by the decision tree for the customer groups, at least the adaptive algorithm is used to dynamically adjust the execution proportion of the collection strategies of each customer group, forming a dynamic feedback process. According to the continuously changing customer characteristics, the execution proportion of the collection strategies of the customer groups divided by customer characteristics is adjusted, realizing the adaptive adjustment of the collection strategies, thus solving the problem that the collection strategies cannot be adaptively adjusted in the prior art.

[0087] In an embodiment of the present application, the generation unit includes an input subunit and a screening subunit. Among them, the input subunit is used to input the historical data of multiple customers in each of the above customer groups into the above decision tree model to obtain multiple initial collection strategies; the screening subunit is used to use a random forest algorithm to screen at least the multiple initial collection strategies to obtain the multiple collection strategies. In this solution, the random forest algorithm can combine the results of multiple decision tree models, reduce the risk of overfitting, and further improve the accuracy and stability of the model, so that the formulation of the collection strategies is more accurate.

[0088] Specifically, according to the decision tree model, by inputting the characteristics of a customer group and following the path from the root node to the leaf node, multiple initial collection strategies for this group can be obtained. The random forest model can generate multiple collection strategies. Each strategy is based on the output of a single decision tree. Through majority voting or weighted average, multiple collection strategies can be generated for each customer group.

[0089] In order to more accurately adjust the execution ratios of multiple collection strategies in the customer group, and thus more accurately adjust the collection strategies, in one embodiment of the present application, the first adjustment unit includes a scoring subunit, a determination subunit, and an adjustment subunit. Among them, the scoring subunit is used to score multiple above-mentioned collection strategies according to the effects of the multiple above-mentioned collection strategies, and obtain multiple scores. The above-mentioned effects include the results of multiple evaluation indicators, and the above-mentioned evaluation indicators at least include the recovery rate and the default rate; the determination subunit is used to determine the optimal collection strategy and the worst collection strategy according to the above-mentioned scores; the adjustment subunit is used to at least increase the first ratio of the above-mentioned optimal collection strategy and decrease the second ratio of the above-mentioned worst collection strategy to obtain the adjusted multiple above-mentioned execution ratios.

[0090] In another embodiment, according to the effects of the multiple above-mentioned collection strategies, the scoring subunit includes a normalization processing module and a scoring module. Among them, the normalization processing module is used to perform normalization processing on the above-mentioned effects of each above-mentioned collection strategy respectively to obtain multiple evaluation index data; the scoring module is used to score each above-mentioned collection strategy according to y = a 1 x 1 +a 2 x 2 +......+a n x n , and obtain multiple above-mentioned scores y. Among them, x 1 ,x 2 ,......x n are multiple evaluation index data, and a 1 ,a 2 ,......a n are the weights corresponding to the multiple above-mentioned evaluation index data. The magnitude of the above-mentioned weights is determined according to the importance degree of the above-mentioned evaluation indicators, and the above-mentioned importance degree is directly proportional to the magnitude of the above-mentioned weights. In this solution, by performing normalization processing on the effects of each collection strategy, the dimensional difference between different evaluation indicators can be further eliminated, enabling them to be compared. Then, weighted summation is performed on each evaluation index data according to the weights to obtain the final score, so that the score can be made more accurate, and thus the collection strategy can be adjusted more accurately and adaptively.

[0091] Specifically, the purpose of normalization is mainly to solve the comparison problems caused by different dimensions and ranges of evaluation metrics. For example, the collection rate may be a percentage, while the collection cost may be measured in monetary units. It is unfair to directly compare the original values of these metrics. Normalization can convert all metrics to a common scale, enabling the data corresponding to the effects to be directly used for comparison between strategies. There are two common normalization methods: Min-Max Scaling, which scales all data to the range of 0 to 1; and Z-score Standardization, which converts the data to a standard normal distribution, i.e., with a mean of 0 and a standard deviation of 1.

[0092] In order to more accurately determine the optimal collection strategy and the worst collection strategy, and thus make the execution ratio of the adjusted collection strategy more accurate, in an embodiment of the present application, the determination subunit includes a first determination module and a second determination module. Among them, the first determination module is used to determine the above-mentioned collection strategy corresponding to the maximum above-mentioned score as the above-mentioned optimal collection strategy; the second determination module is used to determine the above-mentioned collection strategy corresponding to the minimum above-mentioned score as the above-mentioned worst collection strategy.

[0093] In another embodiment, the first adjustment unit further includes a judgment subunit and an execution subunit. Among them, the judgment subunit is used to judge whether the absolute value of the difference between the first score of the above-mentioned optimal collection strategy and the second score of the above-mentioned worst collection strategy is within a preset range; the execution subunit is used to, when the absolute value of the difference between the first score of the above-mentioned optimal collection strategy and the second score of the above-mentioned worst collection strategy is within the above-mentioned preset range, at least adjust the parameters of the above-mentioned clustering algorithm, and repeat the execution of the above-mentioned first partitioning step, the above-mentioned generating step, and the above-mentioned evaluation step at least once until the absolute value of the difference between the first score of the above-mentioned optimal collection strategy and the second score of the above-mentioned worst collection strategy is not within the above-mentioned preset range. In this solution, by continuously adjusting the execution ratio of the optimal collection strategy and the worst collection strategy and adjusting the parameters of the clustering algorithm, the best combination of collection strategies is further searched. By judging whether the absolute value of the score difference between the optimal collection strategy and the worst collection strategy is within the preset range, it can be further determined whether the current strategy combination is effective, so that the execution ratio of the collection strategy is more accurate, and further the collection strategy can be adjusted more accurately.

[0094] Specifically, first, the system calculates the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy. The absolute value of this difference reflects the maximum difference in strategy effectiveness. A preset range is set, such as a score difference of 0.1 to 0.5. If the difference between the optimal strategy and the worst strategy is within this range, it is considered that the current strategy has limited room for optimization, and more detailed customer group segmentation or strategy generation may be required. Secondly, if the score difference between the optimal and worst strategies is within the preset range, the system will re-examine the accuracy of customer group segmentation and optimize the group segmentation by adjusting the parameters of the clustering algorithm. For example, the number of clusters, the initialization method of cluster centers, the distance metric standard, etc. can be adjusted to obtain a more accurate customer group that can better distinguish strategy effectiveness. After that, after adjusting the parameters of the clustering algorithm, the system will re-execute the first segmentation step of customer group segmentation, and then generate a new collection strategy using the decision tree model according to the newly segmented groups, that is, the generation step. Then, the strategy evaluation step is executed, and the results of strategy execution are collected, including indicators such as the repayment rate and default rate, and each strategy is scored again. Finally, the system determines again whether the score difference between the optimal and worst strategies is within the preset range. If not, it means that the strategy effectiveness difference is significant, and the strategy ratio adjustment stage can be entered. If it is still within the preset range, the system will continue to adjust the parameters of the clustering algorithm and repeat the above segmentation, generation, and evaluation steps until the score difference between the optimal and worst strategies exceeds the preset range, that is, the strategy effectiveness difference is significant enough to perform strategy adjustment.

[0095] In order to ensure that the sum of the execution ratios of the optimal collection strategy and the worst collection strategy remains unchanged, thereby making the distribution of collection strategies more balanced, in an embodiment of the present application, the adjustment subunit includes an increase module and a decrease module. The increase module is used to increase the first ratio of the optimal collection strategy by a first preset ratio to obtain an adjusted first execution ratio. The decrease module is used to decrease the second ratio of the worst collection strategy by a second preset ratio to obtain an adjusted second execution ratio. The first preset ratio is equal to the second preset ratio, and the first ratio is equal to the second ratio.

[0096] The above-mentioned collection strategy adjustment device includes a processor and a memory. The above-mentioned acquisition unit, first segmentation unit, generation unit, evaluation unit, first adjustment unit, second segmentation unit, second adjustment unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned each module is located in different processors in any combination form.

[0097] The processor contains a kernel, which retrieves the corresponding program units from the memory. One or more kernels can be set, and by adjusting the kernel parameters, the problem that the collection strategy cannot be adaptively adjusted in the prior art can be solved.

[0098] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

[0099] An embodiment of the present invention provides a computer-readable storage medium, and the computer-readable storage medium includes a stored program. When the program runs, it controls the device where the computer-readable storage medium is located to execute the adjustment method of the collection strategy, and the method includes:

[0100] Step S201, an acquisition step, to acquire the historical data of multiple customers, and the historical data at least includes basic information, income situation, and transaction flow.

[0101] Step S202, a first division step, to divide multiple customers into different customer groups according to the historical data by using a clustering algorithm.

[0102] Step S203, a generation step, to generate multiple collection strategies for each customer group at least by using a decision tree model, and the collection strategy is to use different collection methods every day within a preset time, and the collection methods at least include making calls and sending text messages.

[0103] Step S204, an evaluation step, to execute multiple collection strategies for each customer group to obtain multiple collection results, and use an adaptive algorithm to evaluate the effects of multiple collection results.

[0104] Step S205, a first adjustment step, to adjust at least the proportions of multiple collection strategies in the customer group according to the effects to obtain multiple adjusted execution proportions.

[0105] Step S206, a second division step, to divide the customers in the customer group to obtain multiple customer subgroups, and the proportions between multiple customer subgroups are equal to the execution proportions.

[0106] Step S207, a second adjustment step, to adjust the collection strategy of each customer subgroup to the corresponding collection strategy, the ratio of each customer subgroup to the customer group is a predetermined proportion, and the ratio of the proportion of the collection strategy corresponding to each customer subgroup to the sum of the proportions of all collection strategies is the predetermined proportion.

[0107] An embodiment of the present invention provides a processor, and the method includes:

[0108] Step S201, acquisition step, acquire the historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records.

[0109] Step S202, first division step, according to the above historical data, use the clustering algorithm to divide the multiple customers into different customer groups.

[0110] Step S203, generation step, generate multiple collection strategies for each of the above customer groups at least using a decision tree model, where the collection strategies are to use different collection methods every day within a preset time, and the collection methods at least include making phone calls and sending text messages.

[0111] Step S204, evaluation step, execute the multiple above collection strategies for each of the above customer groups to obtain multiple collection results, and use an adaptive algorithm to evaluate the effectiveness of the multiple above collection results.

[0112] Step S205, first adjustment step, adjust at least according to the above effectiveness the proportions of the multiple above collection strategies in the above customer groups to obtain multiple adjusted execution proportions.

[0113] Step S206, second division step, divide the customers in the above customer groups to obtain multiple customer subgroups, and the proportions between the multiple above customer subgroups are equal to the above execution proportions.

[0114] Step S207, second adjustment step, adjust the above collection strategies of each of the above customer subgroups to the corresponding above collection strategies, the ratio of each of the above customer subgroups to the above customer group is a predetermined proportion, and the ratio of the proportion of the above collection strategies corresponding to each of the above customer subgroups to the sum of the proportions of all the above collection strategies is the above predetermined proportion.

[0115] An embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements at least the following steps:

[0116] Step S201, acquisition step, acquire the historical data of multiple customers, where the historical data at least includes basic information, income situation, and transaction records.

[0117] Step S202, first division step, according to the above historical data, use the clustering algorithm to divide the multiple customers into different customer groups.

[0118] Step S203, generation step, generate multiple collection strategies for each of the above customer groups at least using a decision tree model, where the collection strategies are to use different collection methods every day within a preset time, and the collection methods at least include making phone calls and sending text messages.

[0119] Step S204, Evaluation Step: Execute multiple of the above collection strategies for each of the above customer groups to obtain multiple collection results, and use an adaptive algorithm to evaluate the effectiveness of the multiple collection results.

[0120] Step S205, First Adjustment Step: Adjust the proportions of multiple of the above collection strategies in the above customer groups at least according to the above effectiveness to obtain multiple adjusted execution proportions.

[0121] Step S206, Second Division Step: Divide the customers in the above customer groups to obtain multiple customer subgroups, and the proportions between the multiple customer subgroups are equal to the above execution proportions.

[0122] Step S207, Second Adjustment Step: Adjust the above collection strategies for each of the above customer subgroups to the corresponding above collection strategies. The ratio of each of the above customer subgroups to the above customer group is a predetermined proportion, and the ratio of the proportion of the above collection strategies corresponding to each of the above customer subgroups to the sum of the proportions of all the above collection strategies is the above predetermined proportion.

[0123] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0124] This application also provides a computer program product, which when executed on a data processing device is adapted to execute a program initialized with at least the following method steps:

[0125] Step S201, Acquisition Step: Acquire the historical data of multiple customers, and the above historical data at least includes basic information, income situation, and transaction records.

[0126] Step S202, First Division Step: Divide multiple of the above customers into different customer groups according to the above historical data by using a clustering algorithm.

[0127] Step S203, Generation Step: Generate multiple collection strategies for each of the above customer groups at least by using a decision tree model. The above collection strategies are to use different collection methods every day within a preset time, and the above collection methods at least include making phone calls and sending text messages.

[0128] Step S204, Evaluation Step: Execute multiple of the above collection strategies for each of the above customer groups to obtain multiple collection results, and use an adaptive algorithm to evaluate the effectiveness of the multiple collection results.

[0129] Step S205, First Adjustment Step: Adjust the proportions of multiple of the above collection strategies in the above customer groups at least according to the above effectiveness to obtain multiple adjusted execution proportions.

[0130] Step S206, the second division step, divides the customers in the above customer group to obtain multiple customer subgroups, and the ratio between the multiple above customer subgroups is equal to the above execution ratio.

[0131] Step S207, the second adjustment step, adjusts the above collection strategies of each of the above customer subgroups to the corresponding above collection strategies. The ratio of each of the above customer subgroups to the above customer group is a predetermined proportion, and the ratio of the proportion of the above collection strategies corresponding to each of the above customer subgroups to the sum of the proportions of all the above collection strategies is the above predetermined proportion.

[0132] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. They can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0133] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0134] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one or more flows or multiple flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0135] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction means that implements the functions specified in one or more of the blocks and / or processes, and / or boxes. Figure 1 in one or more of the processes and / or boxes Figure 1 specified in one or more of the boxes and / or processes.

[0136] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes and / or boxes Figure 1 in one or more of the processes and / or boxes Figure 1 specified in one or more of the boxes and / or processes.

[0137] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0138] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read only memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

[0139] Computer-readable media includes both permanent and non-permanent, removable and non-removable media implemented by any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technologies, compact disc read only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0140] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0141] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0142] In the above embodiments, first, historical data of multiple customers is obtained; second, according to the historical data of the customers, multiple customers are divided into different customer groups by using a clustering algorithm; then, at least a decision tree model is used to generate multiple collection strategies for each customer group; the multiple collection strategies are executed, and an adaptive algorithm is used to evaluate the effect of the generated results after executing the collection strategies; according to the effect, the proportion of executing the collection strategies in the customer group is adjusted, and the customers in the customer group are divided according to the proportion to obtain customer subgroups; finally, the collection strategies of each customer subgroup are adjusted to the collection strategies with the same proportion. Compared with the prior art, the solution of the present application uses an adaptive algorithm to monitor in real time the collection execution results of customer groups with different characteristics divided by the clustering algorithm, and according to the collection execution results, dynamically adjusts the proportion of the collection strategies. According to the proportion situation, the collection strategies of the customer group will be gradually adjusted to the collection strategy with the largest proportion, that is, this method uses the clustering algorithm to subdivide customers into different customer groups according to customer characteristics, and according to the results of different collection strategies generated by the decision tree for the customer groups, at least uses the adaptive algorithm to dynamically adjust the execution proportion of the collection strategies of each customer group, forming a dynamic feedback process. According to the continuously changing customer characteristics, the execution proportion of the collection strategies of the customer groups divided by customer characteristics is adjusted, realizing the adaptive adjustment of the collection strategies, thereby solving the problem that the collection strategies cannot be adaptively adjusted in the prior art.

[0143] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A method for adjusting a collection strategy, characterized in that: include: An acquisition step of acquiring historical data of a plurality of customers, wherein the historical data at least includes basic information, income status and transaction flow; A first division step is to divide the plurality of customers into different customer groups using a clustering algorithm according to the historical data; A generating step, at least using a decision tree model to generate a plurality of collection strategies for each customer group, wherein the collection strategy is to use different collection methods every day within a preset time, and the collection methods at least include making phone calls and sending text messages; An evaluation step, executing a plurality of the collection strategies for each of the customer groups to obtain a plurality of collection results, and using an adaptive algorithm to evaluate the effectiveness of the plurality of collection results; A first adjustment step, at least according to the effect, adjusting the ratio of the plurality of collection strategies in the customer group to obtain a plurality of adjusted execution ratios; A second division step is to divide the customers in the customer group into multiple customer subgroups, wherein the ratio between the multiple customer subgroups is equal to the execution ratio; The second adjustment step is to adjust the collection strategy of each customer subgroup to the corresponding collection strategy, the ratio of each customer subgroup to the customer group is a predetermined proportion, and the ratio of the proportion of the collection strategy corresponding to each customer subgroup to the sum of the proportions of all the collection strategies is the predetermined proportion.

2. The method for adjusting the collection strategy according to claim 1, characterized in that: The generation steps include: Inputting the historical data of the plurality of customers in each of the customer groups into the decision tree model to obtain a plurality of initial collection strategies; By using a random forest algorithm, at least a plurality of the initial debt collection strategies are screened to obtain a plurality of the debt collection strategies.

3. The method for adjusting the collection strategy according to claim 1, characterized in that: The first adjustment step includes: Scoring the multiple collection strategies according to their effects to obtain multiple scores, wherein the effects include results of multiple evaluation indicators, and the evaluation indicators include at least a collection rate and a default rate; Determine the best collection strategy and the worst collection strategy based on the scores; At least the first ratio of the optimal debt collection strategy is increased and the second ratio of the worst debt collection strategy is reduced to obtain the adjusted multiple execution ratios.

4. The method for adjusting the collection strategy according to claim 3, characterized in that: According to the effects of the multiple collection strategies, the multiple collection strategies are scored to obtain multiple scores, including: Normalizing the effects of each of the collection strategies to obtain multiple evaluation index data; According to y=a1x1+a2x2+......+a n x n , score each of the collection strategies to obtain multiple scores y, where x1, x2, ... x n are multiple evaluation index data, a1, a2, ... a n are weights corresponding to the plurality of evaluation indicator data.

5. The method for adjusting the collection strategy according to claim 3, characterized in that: Based on the scores, the best and worst collection strategies are determined, including: Determining the collection strategy corresponding to the maximum score as the optimal collection strategy; The collection strategy corresponding to the minimum score is determined as the worst collection strategy.

6. The method for adjusting the collection strategy according to claim 3, characterized in that: Before at least adjusting the first ratio of the optimal debt collection strategy and the second ratio of the worst debt collection strategy to obtain the adjusted execution ratios, the method further includes: Determining whether an absolute value of a difference between a first score of the optimal debt collection strategy and a second score of the worst debt collection strategy is within a preset range; When the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy is within the preset range, at least the parameters of the clustering algorithm are adjusted, and the first division step, the generation step and the evaluation step are repeated at least once until the absolute value of the difference between the first score of the optimal collection strategy and the second score of the worst collection strategy is no longer within the preset range.

7. The method for adjusting the collection strategy according to claim 3, characterized in that: At least adjusting the first ratio of the optimal collection strategy and the second ratio of the worst collection strategy to obtain a plurality of adjusted execution ratios includes: Increasing the first ratio of the optimal collection strategy by a first preset ratio to obtain an adjusted first execution ratio; The second ratio of the worst collection strategy is reduced by a second preset ratio to obtain an adjusted second execution ratio, wherein the first preset ratio is equal to the second preset ratio, and the first ratio is equal to the second ratio.

8. A device for adjusting a collection strategy, characterized in that: include: An acquisition unit, configured to execute the acquisition step to acquire historical data of a plurality of customers, wherein the historical data at least includes basic information, income status and transaction flow; A first division unit, configured to perform a first division step, and divide the plurality of customers into different customer groups using a clustering algorithm according to the historical data; A generating unit, configured to execute the generating step, at least using a decision tree model to generate a plurality of collection strategies for each customer group, wherein the collection strategies are to use different collection methods every day within a preset time, and the collection methods at least include making phone calls and sending text messages; An evaluation unit, configured to execute an evaluation step, execute a plurality of the collection strategies for each of the customer groups, obtain a plurality of collection results, and evaluate the effects of the plurality of collection results using an adaptive algorithm; A first adjustment unit, configured to execute a first adjustment step, and adjust the ratio of the plurality of collection strategies in the customer group at least according to the effect, to obtain a plurality of adjusted execution ratios; A second dividing unit, configured to execute a second dividing step, dividing the customers in the customer group to obtain a plurality of customer subgroups, wherein a ratio between the plurality of customer subgroups is equal to the execution ratio; The second adjustment unit is used to execute the second adjustment step, adjusting the collection strategy of each of the customer subgroups to the corresponding collection strategy, the ratio of each of the customer subgroups to the customer group is a predetermined proportion, and the ratio of the proportion of the collection strategy corresponding to each of the customer subgroups to the sum of the proportions of all the collection strategies is the predetermined proportion.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for adjusting the collection strategy according to any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the method for adjusting the collection strategy described in any one of claims 1 to 7.

11. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include an adjustment method for executing the collection strategy described in any one of claims 1 to 7.

Images