Method and system for allocating financial correspondence cases for special lawyers based on multiple regression model

Abstract

The present invention discloses a method and system for allocating special lawyer financial correspondence based on a multiple regression model. The system includes a client device connected to a distributed database device; the distributed database device is sequentially connected to a telescopic server, a cloud platform device, an intelligent case allocation processing device, a real-time monitoring device, and a terminal visualization large screen. The method of the present invention determines the comprehensive recovery ability score of a customer service based on historical operation data extracted from a distributed database; ranks the customer service's ability to collect payments on overdue cases to obtain a comprehensive ability ranking; finally, the xgboost algorithm is used to score the recovery probability of overdue cases, and according to the recovery probability score value and the comprehensive ability ranking, a multiple regression model algorithm is used to allocate special lawyer financial correspondence, so that high-scoring cases are allocated to customer service with a high comprehensive ability ranking. This greatly improves the efficiency and accuracy of case allocation, and effectively improves the recovery rate of overdue cases.

Description

Technical Field

[0001] The present invention relates to the technical field of financial big data processing, belongs to the category of statistical models and systems, and in particular to a method and system for special lawyer financial correspondence assignment based on a multivariate regression model. Background Art

[0002] With the vigorous development of big data technology, the need for financial credit risk control to shift from traditional manual risk control to digital risk control is becoming more and more urgent. However, in the actual financial case allocation business, the traditional method of allocating cases has long relied on customer service team leaders to allocate cases to customer service staff based on manual experience. Manual experience allocation is not scientific and reasonable, resulting in low case allocation efficiency and low case recovery rate (Deng Songgao; Li Daqu; Ma Sheng; Su Zhixin; Sun Min; Chen Mingping, A system and method for automatic case allocation in court trials, 2016, 1-2).

[0003] However, the current case assignment mechanism also has many problems. Due to the large differences in case type, difficulty, degree of default, lending on different platforms, etc. between different overdue cases, different capabilities are required of customer service personnel. There is currently a lack of scientific methods to accurately assess case workload and customer service workload using the existing business unified platform system data, and it is difficult to achieve the optimal matching of people and cases. The existing case assignment mechanism is "designated case assignment as the main method, and random case assignment as the auxiliary method". The system assigns cases according to the predetermined workflow. There is a lack of scientific and optimal analysis of human-case matching data. It is difficult for personnel to interact with the system for feedback. At the same time, the system cannot discover the potential capability characteristics of new customer service personnel with no history of case assignment (Zhang Qingbing, Case Assignment Method and Device Based on Capability Matching, 2021, 1-2).

[0004] To this end, the present invention establishes a new method and system, which will effectively solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a method and system for special lawyer financial letter assignment based on a multivariate regression model, aiming to solve the technical problem that due to the large differences in case types, difficulty, degree of default, lending on different platforms, etc. among different overdue cases, the existing case assignment mechanism relies on customer service team leaders to assign cases based on manual experience, which makes it difficult to achieve the optimal matching of people and cases.

[0006] To achieve the above-mentioned purpose, the present invention proposes a method for assigning special lawyer financial letters based on a multiple regression model. The method is implemented based on a system for assigning special lawyer financial letters based on a multiple regression model. The method includes: extracting the historical operation data of each customer service from a distributed database, and the distributed database is used to distribute the case information of special lawyer financial letters; determining the comprehensive recovery ability score of each customer service according to the historical operation data; ranking the ability of each customer service to collect overdue cases according to the comprehensive recovery ability score, and obtaining a comprehensive ability ranking; using the xgboost algorithm to score the recovery probability of the overdue case, and obtaining a recovery probability score value; according to the comprehensive ability ranking and the recovery probability score value of each customer service, using the multiple regression model algorithm to assign cases to the special lawyer financial letters.

[0007] The present invention can significantly improve the efficiency and accuracy of case assignment through an intelligent case assignment mechanism, improve the efficiency of recovering overdue cases, and maximize the case recovery rate.

[0008] To achieve the above purpose, the technical solution provided by the present invention is as follows:

[0009] Step 1: System construction.

[0010] The special lawyer financial correspondence distribution system based on the multivariate regression model consists of the following key modules:

[0011] A client device, the client device is connected to the distributed database device, and the client device is used to receive and process case information;

[0012] The distributed database device is connected to the telescopic server, cloud platform device, intelligent case processing device, real-time monitoring device, and terminal visualization large screen in sequence;

[0013] The distributed database device is used to store data on multiple physical nodes and implement data management through distributed computing and storage technology. It has high scalability and fault tolerance and supports large-scale data and high-concurrency access requests.

[0014] The telescopic server and the cloud platform device are used to provide elastic computing and storage capabilities and support dynamic expansion and processing of data;

[0015] The intelligent case assignment processing device implements a special lawyer financial correspondence case assignment method based on a multivariate regression model to extract data, calculate customer service capabilities, and perform case assignment;

[0016] The real-time monitoring equipment is used to track the payment results and case assignment of special lawyer financial letters to ensure that the cases are completed reasonably and on time;

[0017] The terminal visualization large screen is used to output customer service and case allocation detailed reports. The customer service and case allocation detailed reports record the allocation and processing process of each special lawyer's financial correspondence, and can display case allocation details, collection effects and strategy evaluation results.

[0018] The entire system builds an efficient and intelligent case distribution platform by integrating distributed databases, scalable servers, cloud platform equipment, intelligent case distribution processing equipment, real-time monitoring equipment and terminal visualization large screens.

[0019] Step 2: Data extraction and cleaning.

[0020] The step of extracting the historical operation data of each customer service from the distributed database includes: extracting the historical operation data of each customer service from the distributed database, the historical operation data including historical collection data and characteristic data of overdue cases; the historical collection data including the average recovery rate, the collection amount and the recovery rate of each aging period; the characteristic data of the overdue cases including repeated collection, the borrower's gender and the common debt situation; after extracting the historical operation data, performing data cleaning and processing to remove outliers and noise data to ensure data quality.

[0021] Step 3: Comprehensive recycling ability score of customer service.

[0022] The step of determining the comprehensive recovery capability score of each customer service representative based on the historical operation data includes: using a weighted average method to assign different weights according to the customer's call volume, effective call volume and recovery rate in historical telephone operations; and determining the comprehensive recovery capability score of each customer service representative based on the different weights and the historical collection data.

[0023] Specifically, in order to evaluate the comprehensive recycling ability score of each customer service representative, the specific formula for calculating the comprehensive recycling ability score using the weighted average method is as follows:

[0024]

[0025] In the formula, x i Represents the customer service's scores on different indicators, including the number of calls, effective number of calls, historical recovery rate, etc. i Represents the weights of these indicators; through this scoring mechanism, the historical performance of customer service is quantified into a comprehensive recovery ability score, which is used to rank all customer service staff.

[0026] Step 4: Case recovery probability scoring.

[0027] The step of scoring the recovery probability of the overdue cases by the xgboost algorithm to obtain the recovery probability score value comprises: binning the basic features of the overdue cases, wherein the basic features include gender, age and number of overdue days; selecting the feature variables with the highest correlation with the recovery effect among the basic features; constructing an xgboost model by the xgboost algorithm, and inputting the feature variables into the xgboost model for training; evaluating the accuracy between the model prediction value and the actual value according to the model effect index, and predicting the recovery probability of the overdue cases of the new sample when the accuracy rate meets the standard; scoring the recovery probability to obtain the recovery probability score value.

[0028] Specifically, the steps of the xgboost algorithm to build the xgboost model include:

[0029] The first step is model training: input the selected feature variables into the xgboost model for training; then perform prediction and evaluation: use model effect indicators such as AUC and KS to evaluate the accuracy of the model and predict the recovery probability of new samples.

[0030] Step 5: Multiple regression model case allocation.

[0031] The step of allocating cases to the special lawyer financial letters using a multiple regression model algorithm based on the comprehensive ability ranking and the recovery probability score value of each customer service includes: analyzing the special lawyer financial letters using a multiple regression model algorithm based on the comprehensive ability ranking and the recovery probability score value of each customer service to determine high-scoring cases and low-scoring cases; allocating the high-scoring cases to customer services with a high comprehensive ability ranking; and allocating the low-scoring cases to other customer services using cluster analysis.

[0032] Specifically, the formula of the multiple regression model algorithm is as follows:

[0033] Y=β0+β1X1+…+β n X n +∈;

[0034] In the formula, Y is the case allocation result, X1, X n are variables that affect case allocation (including customer service comprehensive recovery ability score and recovery probability score), β0, β1, …, β n is the regression coefficient, and ∈ is the error term. By optimizing the model parameters, high-scoring cases are assigned to customer service staff with high comprehensive ability rankings, and low-scoring cases are matched with the most suitable customer service staff using cluster analysis.

[0035] Step 6. Real-time monitoring and feedback.

[0036] Track the collection results and case assignments through real-time monitoring equipment to ensure that each case is assigned reasonably and completed on time. At the same time, the terminal visual large screen outputs detailed reports on customer service and case assignments, records the assignment and processing process, analyzes the collection effect, evaluates the effectiveness of the case assignment strategy, and provides data support for future strategy adjustments.

[0037] One or more technical solutions proposed in the present invention have at least the following technical effects: the system of the present invention includes a client device, and the client device is connected to a distributed database device; the distributed database device is connected to a telescopic server, a cloud platform device, an intelligent case processing device, a real-time monitoring device, and a terminal visualization large screen in turn. The method of the present invention extracts the historical operation data of each customer service from a distributed database, and the distributed database is used to distribute the case information of special lawyer financial letters; according to the historical operation data, the comprehensive recovery ability score of each customer service is determined; according to the comprehensive recovery ability score, the ability of each customer service to collect overdue cases is ranked to obtain a comprehensive ability ranking; the recovery probability of the overdue case is scored using the xgboost algorithm to obtain a recovery probability score value; according to the comprehensive ability ranking and the recovery probability score value of each customer service, a multivariate regression model algorithm is used to allocate cases to the special lawyer financial letters. Since the present invention adopts the xgboost algorithm to score the recovery probability of overdue cases, and adopts the multivariate regression model algorithm to allocate cases according to the comprehensive ability ranking of customer service personnel and the recovery probability score value of overdue cases, cases with high scores are allocated to customer service personnel with high comprehensive ability ranking, thereby realizing intelligent case allocation in the big data system, effectively solving the problem that the traditional method has long relied on customer service team leaders to allocate cases based on manual experience, effectively improving the recovery rate of overdue cases, achieving the goal of intelligent case allocation in a big data environment, and greatly improving the efficiency and accuracy of case allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0040] Figure 1 It is a flow chart of the method for classifying special lawyer financial correspondence based on the multivariate regression model of the present invention;

[0041] Figure 2It is a structural schematic diagram of the special lawyer financial correspondence distribution system based on the multivariate regression model of the present invention.

[0042] Description of Figure Numbers:

[0043] 1. Client device; 2. Distributed database device; 3. Retractable server; 4. Intelligent case processing device; 5. Real-time monitoring device; 6. Terminal visualization large screen.

[0044] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0045] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0047] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, close, far...) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0048] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0049] The embodiment of the present invention provides a method for allocating special lawyer financial correspondence cases based on a multivariate regression model.

[0050] It should be noted that the executor of this embodiment can be a computing service device with data extraction, recovery probability scoring and case allocation functions, such as a personal computer, server, etc., or an electronic device that can realize the above functions, a special lawyer financial letter assignment device based on a multiple regression model that executes the special lawyer financial letter assignment method based on a multiple regression model of the present invention, etc. This embodiment does not limit this.

[0051] In this embodiment, the method for assigning special lawyer financial letters based on a multivariate regression model includes: extracting historical operation data of each customer service from a distributed database, the distributed database being used to distribute the case information of special lawyer financial letters; determining the comprehensive recovery ability score of each customer service based on the historical operation data; ranking each customer service's ability to collect payments on overdue cases based on the comprehensive recovery ability score to obtain a comprehensive ability ranking; using the xgboost algorithm to score the recovery probability of the overdue cases to obtain a recovery probability score value; and using a multivariate regression model algorithm to assign cases to the special lawyer financial letters based on the comprehensive ability ranking and the recovery probability score value of each customer service.

[0052] It should be noted that a distributed database is a database system that stores data in multiple different physical locations and is used for distributed storage of case information of special attorney financial correspondence.

[0053] Among them, special lawyer financial letters refer to letters related to the financial field sent by lawyers to relevant parties on behalf of their clients or based on specific legal matters. Special lawyer financial letters include overdue cases with different sample characteristics.

[0054] It should be noted that the historical operation data is the historical operation data of the customer service for overdue cases, including the repayment amount of overdue cases, the recovery rate of each account age period of overdue cases, the gender of borrowers in overdue cases, the common debt situation of overdue cases, the number of calls, the effective number of calls, the historical recovery rate, etc.

[0055] It should be noted that the comprehensive recovery ability score is a quantitative indicator used to measure customer service’s ability to collect payments on overdue cases.

[0056] Among them, overdue cases refer to special lawyer financial letters corresponding to the failure to perform contractual obligations as agreed within the specified time. Collection ability refers to the customer service's ability to collect payments on overdue cases.

[0057] It should be noted that the xgboost algorithm is a machine learning algorithm based on the gradient boosting tree, which can be used to predict the recovery probability of overdue cases.

[0058] The recovery probability refers to the probability of successful collection of overdue cases by customer service. The xgboost algorithm can be used to predict the probability of successful collection of new overdue cases by customer service (i.e., the recovery probability score value).

[0059] It should be noted that the multiple regression model algorithm is a statistical analysis method used to measure the customer service's ability to resolve overdue cases. Through the multiple regression model algorithm, high-scoring cases can be assigned to customer service staff with high comprehensive ability rankings, and low-scoring cases can be matched to other most suitable customer service staff.

[0060] In this implementation, the xgboost algorithm is used to score the recovery probability of overdue cases. According to the comprehensive capabilities of customer service personnel and the recovery probability score values ​​of overdue cases, a multivariate regression model algorithm is used to allocate cases, so that high-scoring cases are allocated to customer service personnel with high comprehensive capability rankings, to ensure that high-scoring cases are efficiently recovered.

[0061] Furthermore, the step of extracting the historical operation data of each customer service from the distributed database includes: extracting the historical operation data of each customer service from the distributed database, the historical operation data including historical collection data and characteristic data of overdue cases; the historical collection data including the average recovery rate, the collection amount and the recovery rate of each aging period; the characteristic data of the overdue cases including repeated collection, borrower gender and common debt situation.

[0062] It should be noted that the average recovery rate is the recovery ratio obtained by dividing the number of cases successfully recovered by the customer service by the total case business volume. The amount of funds recovered is the amount of funds actually recovered by the financial letter. The recovery rate of each aging period is the recovery rate of different aging time periods (such as aging within 30 days, aging 30-60 days, etc.).

[0063] It should be noted that repeated collection refers to the situation where the customer service collects the debt of the same financial letter multiple times, including the number and frequency of collections, etc. The common debt situation refers to the situation where the borrower of the financial letter has debts in other places in addition to the current debt involved.

[0064] In this embodiment, a distributed database is used, which has higher scalability and fault tolerance than a traditional stand-alone database, and can handle large-scale data and high-concurrency access requests. A distributed database can be easily expanded to cope with the growth of data scale by adding nodes without the need to reconstruct the entire system. It usually has data redundancy and failover capabilities, so that data is still available even if some nodes fail. A distributed database can distribute data across multiple nodes, thereby reducing the load on a single node and improving query performance.

[0065] Furthermore, the step of determining the comprehensive recovery capability score of each customer service representative based on the historical operation data includes: using a weighted average method to assign different weights according to the number of calls, effective number of calls and recovery rate of the customer in historical telephone operations; and determining the comprehensive recovery capability score of each customer service representative based on the different weights and the historical collection data.

[0066] It should be noted that the weighted average method is a method of calculating the average value based on the importance (ie, weight) of the data of historical telephone operations.

[0067] Specifically, in order to evaluate the comprehensive recycling ability score of each customer service representative, the specific formula for calculating the comprehensive recycling ability score using the weighted average method is as follows:

[0068]

[0069] In the formula, x i Represents the customer service's scores on different indicators, including the number of calls, effective number of calls, historical recovery rate, etc. i Represents the weights of these indicators; through this scoring mechanism, the historical performance of customer service is quantified into a comprehensive recovery ability score, which is used to rank all customer service staff.

[0070] In this embodiment, the comprehensive recycling ability score of each customer service representative can be calculated by weighted average method.

[0071] Furthermore, the step of scoring the recovery probability of the overdue cases with the xgboost algorithm to obtain a recovery probability score value includes: binning the basic characteristics of the overdue cases, wherein the basic characteristics include gender, age and number of overdue days; selecting the characteristic variables with the highest correlation with the recovery effect among the basic characteristics; constructing an xgboost model with the xgboost algorithm, and inputting the characteristic variables into the xgboost model for training; evaluating the accuracy between the model prediction value and the actual value according to the model effect index, and predicting the recovery probability of the overdue cases of the new sample when the accuracy rate meets the standard; scoring the recovery probability to obtain a recovery probability score value.

[0072] It should be noted that the xgboost model is a boosted tree model built using the xgboost algorithm, which is used to fit data and predict the probability of recovery of overdue cases.

[0073] Among them, the model effect index is a quantitative indicator used to evaluate and measure the performance of the xgboost model, such as AUC (Area Under the Curve), KS value, etc., which are used to evaluate model performance.

[0074] In this implementation, the xgboost algorithm can be used to accurately predict the probability of recovery of overdue cases to provide support for subsequent case allocation.

[0075] Furthermore, the step of allocating cases to the special lawyer financial letters using a multiple regression model algorithm based on the comprehensive ability ranking and the recovery probability score value of each customer service includes: analyzing the special lawyer financial letters using a multiple regression model algorithm based on the comprehensive ability ranking and the recovery probability score value of each customer service to determine high-scoring cases and low-scoring cases; allocating the high-scoring cases to customer services with high comprehensive ability rankings; and allocating the low-scoring cases to other customer services using cluster analysis.

[0076] It should be noted that cluster analysis is a method of classifying low-scoring cases according to their similarities or distances with high concentration, so as to assign them to the most suitable customer service representative.

[0077] Specifically, the formula of the multiple regression model algorithm is as follows:

[0078] Y=β0+β1X1+…+β n X n +∈;

[0079] In the formula, Y is the case allocation result, X1, X n are variables that affect case allocation (including customer service comprehensive recovery ability score and recovery probability score), β0, β1, …, β n is the regression coefficient, and ∈ is the error term. By optimizing the model parameters, high-scoring cases are assigned to customer service staff with high comprehensive ability rankings, and low-scoring cases are matched with the most suitable customer service staff using cluster analysis.

[0080] In this implementation, high-scoring cases are assigned to customer service staff with high comprehensive ability rankings, while low-scoring cases are analyzed using cluster analysis to analyze the high concentration of historical low-scoring cases in the hands of corresponding customers, so that low-scoring cases are assigned to the most suitable customer service staff, thereby improving the overall case recovery rate.

[0081] In the technical solution provided in this embodiment, the intelligent case division of the big data system is realized, which effectively solves the problem that the traditional method relies on the customer service team leader to divide the case based on manual experience for a long time. The efficiency and accuracy of case division are improved, so that the case recovery rate is maximized.

[0082] In order to help those skilled in the art better understand the technical solution of the present invention, a specific example will be used to demonstrate the calculation and analysis of the entire process, including the comprehensive recovery ability score of customer service, the recovery probability score of overdue cases, and the case allocation of the multivariate regression model. The following are detailed steps and calculation processes, but they are not limited to this solution. Figure 1 , Figure 1 The flowchart of the method for classifying special lawyer financial correspondence based on the multivariate regression model of the present invention is as follows:

[0083] Step 1: Get the historical operation data of each customer service representative from the distributed database.

[0084] Assume that there are three customer service representatives (Customer Service A, Customer Service B, and Customer Service C) and three overdue cases (Case 1, Case 2, and Case 3). These cases need to be assigned to the customer service representatives to maximize the recovery rate. For example, the historical telephone operation data of the customer service representatives is shown in Table 1:

[0085] Table 1: Historical average recovery rate data table of customer service

[0086] customer service Call volume Effective number of calls Average recovery rate Customer Service A 100 80 0.75 Customer Service B 120 70 0.70 customer service 90 85 0.80

[0087] For example, the basic characteristics of overdue cases are shown in Table 2:

[0088] Table 2: Characteristics of overdue cases

[0089] case gender age Expected number of days Case 1 male 35 60 Case 2 female 28 45 Case 3 male 40 90

[0090] At this time, it is necessary to clean the above data of outliers and noise data.

[0091] Step 2: Calculate the customer service comprehensive recycling ability score.

[0092] After cleaning the outliers and noise data, the weighted average method is used to calculate the comprehensive recycling ability score of each customer service representative. At this time, the weight is assumed to be:

[0093] The number of incoming calls w1 = 0.3;

[0094] The effective number of incoming calls w2 = 0.3;

[0095] Average recovery rate w3 = 0.4;

[0096] Then, according to the comprehensive recycling capacity score = w1x1+w2x2+w3x3, calculate:

[0097] Customer service A's comprehensive recycling ability score = 0.3*100+0.3*80+0.4*0.75=30+24+30=84.3;

[0098] Customer service B's comprehensive recycling ability score = 0.3*120+0.3*70+0.4*0.7=36+21+28=85;

[0099] Customer service C’s comprehensive recycling capability score = 0.3*90+0.3*85+0.4*0.8=27+25.5+32=84.5.

[0100] In summary, the comprehensive capabilities of customer service are ranked as follows:

[0101] Customer Service B: 85, Customer Service C: 84.5, Customer Service A: 84.3.

[0102] Step 3: Use the xgboost algorithm to score the case recovery probability.

[0103] The xgboost algorithm is used to calculate the recovery probability of each case. Assume that after training the xgboost model, the recovery probability score values ​​are shown in Table 3:

[0104] Table 3: Recovery probability scoring table

[0105] case Recovery probability score Case 1 0.80 Case 2 0.60 Case 3 0.75

[0106] Step 4: Use a multivariate regression model to match cases with customer service.

[0107] The following multiple regression model was used for case allocation:

[0108] Allocation score = β0 + β1 * customer service comprehensive ability score + β2 * recovery probability score value;

[0109] Assume that the regression coefficients of the model are:

[0110] β0 = (bias term is zero, simplified processing);

[0111] β1 = 0.6;

[0112] β2 = 0.4;

[0113] At this point, calculate the distribution score of each case under each customer service:

[0114] Assigned scores for Case 1:

[0115] Customer Service A: 0.6*84.3+0.4*0.80=50.58+0.32=50.9;

[0116] Customer Service B: 0.6*85+0.4*0.80=51+0.32=51.32;

[0117] Customer Service C: 0.6*84.5+0.4*0.80=50.7+0.32=51.02.

[0118] Assigned scores for Case 2:

[0119] Customer Service A: 0.6*84.3+0.4*0.60=50.58+0.24=50.82;

[0120] Customer Service B: 0.6*85+0.4*0.60=51+0.24=51.24;

[0121] Customer Service C: 0.6*84.5+0.4*0.60=50.7+0.24=50.94.

[0122] Assigned scores for Case 3:

[0123] Customer Service A: 0.6*84.3+0.4*0.75=50.58+0.3=50.88;

[0124] Customer Service B: 0.6*85+0.4*0.75=51+0.3=51.3;

[0125] Customer Service C: 0.6*84.5+0.4*0.75=50.7+0.3=51.

[0126] Step 5: Case division results.

[0127] Based on the above allocated scores, high-scoring cases are concentrated in customer service staff with high comprehensive recovery ability scores, and cluster analysis is used to optimally match the characteristics of low-scoring cases with the remaining customer service staff.

[0128] Based on the above distribution scores, the distribution is as follows:

[0129] Case 1: Assigned to Customer Service B (51.32 points);

[0130] Case 2: Assigned to Customer Service B (51.24 points);

[0131] Case 3: Assigned to Customer Service B (51.3 points).

[0132] Step 6: Use real-time monitoring equipment to monitor whether cases are allocated as required and output the actual collection effect.

[0133] In this example, all cases are assigned to customer service B with the highest comprehensive ability, because customer service B has an advantage in the combined score of comprehensive ability and case recovery probability. This shows that using a multivariate regression model to allocate cases can effectively allocate cases with high recovery probabilities to the most suitable customer service, thereby maximizing recovery efficiency. Compared with traditional empirical allocation methods, it can more accurately match customer service and cases, improve the overall recovery rate, and through real-time data monitoring, it can dynamically adjust strategies to further optimize allocation effects.

[0134] The above-mentioned special lawyer financial correspondence assignment method based on the multiple regression model is implemented based on the special lawyer financial correspondence assignment system based on the multiple regression model proposed in this embodiment. Figure 2 , Figure 2It is a structural schematic diagram of the special lawyer financial correspondence distribution system based on the multivariate regression model of the present invention.

[0135] The special lawyer financial correspondence assignment system based on the multivariate regression model described in this embodiment is implemented by the special lawyer financial correspondence assignment method based on the multivariate regression model as described above, and the system includes:

[0136] A client device 1, wherein the client device 1 is connected to a distributed database device 2;

[0137] The distributed database device 2 is connected to the telescopic server 3, the cloud platform device, the intelligent case processing device 4, the real-time monitoring device 5, and the terminal visualization large screen 6 in sequence;

[0138] The real-time monitoring device 5 is used to track the payment results and case distribution of special lawyer financial letters;

[0139] The terminal visualization large screen 6 is used to output a detailed report on customer service and case allocation, which records the allocation and processing process of each special lawyer's financial correspondence.

[0140] Furthermore, the distributed database device 2 is used to store data on multiple physical nodes and implement data management through distributed computing and storage technology.

[0141] It should be noted that distributed database devices are used to store data on multiple physical nodes and realize data management through distributed computing and storage technologies. They have high scalability and fault tolerance, and support large-scale data and high-concurrency access requests.

[0142] Scalable servers and cloud platform equipment are used to provide elastic computing and storage capabilities and support dynamic expansion and processing of data.

[0143] The intelligent case assignment processing equipment executes the above-mentioned special lawyer financial correspondence assignment method based on the multivariate regression model to extract data, calculate customer service capabilities, and assign cases.

[0144] Real-time monitoring equipment is used to track the progress of payment collection and the assignment of each case, so that the assignment of cases can be completed on time, as required, and reasonably. The terminal visualization screen can output detailed customer service capabilities and case assignment details, recording the assignment and processing process of each case. Analyze the payment collection effect, evaluate the effectiveness of the case assignment strategy, and provide data support for future strategy adjustments.

[0145] In this embodiment, the entire system builds an efficient and intelligent case distribution platform by integrating with distributed databases, scalable servers, cloud platform equipment, intelligent case distribution processing equipment, real-time monitoring equipment and terminal visualization large screens.

[0146] It should be noted that the above examples are only used to understand the present invention and do not constitute a limitation on the special lawyer financial letter division method and system based on the multivariate regression model of the present invention. More simple transformations based on this technical concept are all within the scope of protection of the present invention.

[0147] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0148] The above descriptions are only some embodiments of the present invention, and are not intended to limit the patent scope of the present invention. All equivalent structural changes made using the contents of the present invention's specification and drawings under the technical concept of the present invention, or directly / indirectly applied in other related technical fields, are included in the patent protection scope of the present invention.

Claims

1. A method for classifying special lawyer financial correspondence based on a multivariate regression model, characterized in that: The method includes: Extracting historical operation data of each customer service from a distributed database, wherein the distributed database is used for distributed storage of case information of special lawyer financial correspondence; Determine the comprehensive recycling capability score of each customer service representative based on the historical operation data; According to the comprehensive collection ability score, the collection ability of each customer service representative on overdue cases is ranked to obtain a comprehensive ability ranking; The xgboost algorithm is used to score the recovery probability of the overdue case to obtain a recovery probability score value; According to the comprehensive ability ranking and the recovery probability score value of each customer service, a multivariate regression model algorithm is used to allocate cases for the special lawyer financial letters.

2. The method according to claim 1, characterized in that The step of extracting the historical operation data of each customer service from the distributed database includes: Extracting historical operation data of each customer service representative from a distributed database, wherein the historical operation data includes historical payment collection data and characteristic data of overdue cases; The historical collection data includes the average collection rate, collection amount and collection rate of each aging period; The characteristic data of overdue cases include repeated collection, borrower’s gender and joint debt situation.

3. The method according to claim 1, characterized in that The step of determining the comprehensive recycling capability score of each customer service representative based on the historical operation data includes: Use the weighted average method to assign different weights to customers based on their historical call volume, effective call volume, and recovery rate; Based on the different weights and the historical collection data, the comprehensive collection ability score of each customer service representative is determined.

4. The method according to claim 1, characterized in that The step of using the xgboost algorithm to score the recovery probability of the overdue case to obtain the recovery probability score value includes: The basic characteristics of the overdue cases are divided into boxes, wherein the basic characteristics include gender, age and number of overdue days; Select the characteristic variable with the highest correlation with the recycling effect among the basic characteristics; Using the xgboost algorithm to build an xgboost model, and inputting the feature variables into the xgboost model for training; The accuracy between the model prediction value and the actual value is evaluated based on the model effect index, and the recovery probability of overdue cases in new samples is predicted when the accuracy reaches the target; The recovery probability is scored to obtain a recovery probability score value.

5. The method according to any one of claims 1 to 4, characterized in that The step of using a multivariate regression model algorithm to allocate cases to the special lawyer financial letters according to the comprehensive ability ranking of each customer service and the recovery probability score value includes: According to the comprehensive ability ranking of each customer service and the recovery probability score value, a multivariate regression model algorithm is used to analyze the special lawyer financial letters to determine high-scoring cases and low-scoring cases; Allocate the high-scoring cases to customer service representatives who rank high in comprehensive ability; Cluster analysis method is used to assign the low-scoring cases to other customer service staff.

6. A special lawyer financial correspondence assignment system based on a multivariate regression model, characterized in that: The special lawyer financial correspondence assignment system based on the multiple regression model is implemented by the special lawyer financial correspondence assignment method based on the multiple regression model as claimed in claim 1, and the system includes: A client device, wherein the client device is connected to the distributed database device; The distributed database device is connected to the telescopic server, cloud platform device, intelligent case processing device, real-time monitoring device, and terminal visualization large screen in sequence; The real-time monitoring equipment is used to track the payment results and case assignment of special lawyer financial letters; The terminal visualization large screen is used to output customer service and case allocation detailed reports, which record the allocation and processing process of each special lawyer's financial correspondence.

7. The system according to claim 6, characterized in that The distributed database device is used to store data on multiple physical nodes and implement data management through distributed computing and storage technology.

Images