CN-121981748-A - Passenger group accurate marketing evaluation classification method for data element analysis

CN121981748ACN 121981748 ACN121981748 ACN 121981748ACN-121981748-A

Abstract

The invention relates to the technical field of bank marketing, and discloses a customer group accurate marketing evaluation classification method for data element analysis, aiming at solving the problems of insufficient pertinence and blindness in adjustment of interest rate of traditional marketing. The method comprises the steps of periodically collecting credit reporting of banking clients according to quarters, extracting information to construct client credit element information, completing customer group classification of the banking clients and marketing grade classification of the banking clients based on the client credit element information, simultaneously carrying out sensitivity classification in combination with market synchronization reference interest rate to determine the interest rate sensitivity grade of each banking client, matching basic differentiated marketing schemes through a rule engine according to the sensitivity grade, obtaining optimal interest rate adjustment amplitude through a Monte Carlo simulation method, integrating the basic differentiated marketing schemes and the basic differentiated marketing schemes into personalized marketing schemes, and pushing the personalized marketing schemes to the banking clients. The invention realizes the accurate positioning and the scientific adjustment of the interest rate of the customer group of the bank customer, and improves the marketing conversion rate and the profit of the business information.

Inventors

LI SHIBING
REN HAO
LIU ZHANZHU
ZHAO SHUANG
HAN YUWEI

Assignees

吉林省通联信用服务有限公司

Dates

Publication Date: 20260505
Application Date: 20251225

Claims (9)

1. A customer group accurate marketing evaluation classification method for data element analysis is characterized by comprising the following steps: S1, collecting credit report of banking client, and carrying out content identification and extraction on the credit report to construct client credit element information of banking client; s2, classifying the stock customers based on the credit factor information of the customers to obtain a plurality of customer group groups, wherein the customer group groups are associated with a single customer group label; S3, based on the credit element information of the customers, carrying out sensitivity classification on the banking customers by combining with the market contemporaneous reference interest rate to obtain the interest rate sensitivity level of each banking customer; s4, calculating the comprehensive score of the banking client based on the client credit element information, and dividing the marketing grade of the banking client according to the comprehensive score; S5, matching a basic differentiated marketing scheme through a rule engine based on customer group labels, marketing grades and interest rate sensitivity grades of the banking clients, obtaining optimal interest rate adjustment amplitude of the banking clients through multi-scene interest rate adjustment simulation, integrating the basic differentiated marketing scheme with the optimal interest rate adjustment amplitude, and generating a personalized marketing scheme of the banking clients; s6, pushing the personalized marketing scheme to the banking stock client through the APP, periodically acquiring credit investigation reports of the banking stock client according to quarters, and carrying out S1-S5 again.
2. The method for classifying customer base accurate marketing assessment of data element analysis according to claim 1, wherein the content recognition extraction extracts credit text content in the credit report by an optical character recognition technology, and adopts a pre-trained financial field NLP model to extract customer base information, loan records and risk related information.
3. The method of claim 2, wherein the customer credit factor information includes customer base information including customer ID, age, occupation, area of interest and industry of interest, and loan record including overdue times, current total liabilities and liabilities rates, and risk related information including loan organization name, loan amount, execution interest rate, loan deadline, loan type, repayment mode and repayment status.
4. The method for classifying a customer base with accurate marketing assessment for data element analysis according to claim 3, wherein the method for classifying the customer base is as follows: s21, selecting initial features based on the credit factor information of the clients, sorting the feature importance of the initial features through LightGBM, and selecting the initial features with the importance ratio more than or equal to 2% and the accumulated importance ratio more than 80% as core features; s22, encoding the category core features in the core features into numerical values, and executing a K-means clustering algorithm on the encoded core features to obtain an initial guest group clustering result; S23, based on the business rules, carrying out rule correction on the initial guest group result through a decision tree model to generate a plurality of guest group groups.
5. The method for classifying a customer base accurate marketing assessment according to claim 2, wherein the method for classifying sensitivity is as follows: S31, based on the credit element information of the clients and the market synchronization reference interest rate, calculating the fluctuation range of the market synchronization reference interest rate in a statistical mode Competitive coefficient of Benzhi Rate Poor interest rate of banking stock customers Average loan remaining term Life cycle of customer loan And loan amount ratio ; S32, will 、、、 And Substituting the Logistic regression model to calculate the basic sensitivity probability, and combining And the industry of banking stock clients, correcting the basic sensitivity probability to obtain the interest rate sensitivity probability; S33, dividing the banking clients into a high interest rate sensitivity level, a medium interest rate sensitivity level, a low interest rate sensitivity level and an interest rate insensitivity level according to the interest rate sensitivity probability.
6. The method for classifying customer base accurate marketing assessment of data element resolution of claim 2, wherein the composite score is obtained by weighting and summing the customer historical performance, customer risk level and customer value potential according to a predetermined rule.
7. The method for classifying the customer base accurate marketing assessment of the data element analysis according to claim 6, wherein the rule engine calculates and integrates the product reference interest rate and the optimal interest rate adjustment range of the basic differentiated marketing scheme based on the customer base differentiated marketing scheme in a preset bank marketing product scheme library matched with the marketing grade, and the personalized marketing scheme of the banking customers is obtained.
8. The method for classifying customer base accurate marketing assessment of data element analysis according to claim 7, wherein the multi-scene interest rate adjustment simulation simulates the customer conversion probability and the interest difference income corresponding to the basic differentiated marketing scheme under each interest rate adjustment scene by a Monte Carlo simulation method, and takes the optimal balance point of the customer conversion probability and the interest difference income as the optimal interest rate adjustment amplitude.
9. The method for classifying customer base accurate marketing assessment according to claim 2, wherein when the interest rate sensitivity level of the banking customer is the interest rate insensitivity level, the basic differentiated marketing scheme of the banking customer is used as the personalized marketing scheme.

Description

Passenger group accurate marketing evaluation classification method for data element analysis Technical Field The invention relates to the technical field of bank marketing, in particular to a customer group accurate marketing evaluation classification method for data element analysis. Background In the technical field of bank marketing, accurate classification of customer groups is a core foundation for realizing differentiated marketing and improving marketing efficiency and poor interest, and core logic of the accurate classification is to construct a client layering system by mining data characteristics such as client attributes, credit behaviors and the like, so as to match an adaptive marketing scheme, and solve the problems of resource waste and insufficient pertinence in traditional marketing. Along with the aggravation of financial market competition and the diversification of customer demands, banks have higher requirements on the classification accuracy of customer groups and the individuation of marketing schemes, and the realization of transformation from broad-spectrum marketing to accurate touchdown is urgently needed by relying on data element analysis and algorithm model optimization. As the patent number publication number CN114997925A discloses a method and a device for classifying the customer groups of the bank, the method constructs a customer attribute vector by collecting the information of the whole customers of the bank, and completes the automatic classification of the customer groups by combining weight distribution and matrix decomposition, thereby breaking through the limitations of traditional manual screening and single dimension division, realizing the automation and scale of the classification of the customer groups and solving the basic problem of high-efficiency grouping of massive customers of the bank. However, the prior art related to bank customer group marketing still has a plurality of remarkable defects that firstly, customer data is not fully extracted and utilized, basic attribute data are depended, core credit elements such as loan records and risk states in credit reporting are not fully mined, and a standardized information extraction mechanism is lacking, so that classified data is on one side, secondly, a customer group classification algorithm is single, closed loop flows of feature screening, clustering and business rule correction are not established, the customer group classification algorithm is easily interfered by redundant features, classification results and actual business scene suitability are poor, customer group labels lack uniqueness and pertinence, thirdly, marketing schemes are not formulated by combining market interest rate dynamics and customer interest rate sensitivity characteristics, interest rate adjustment depends on experience judgment, blindness is strong, customer conversion and poor income of banking information are difficult to balance, and fourthly, a dynamic optimization mechanism is lacking, customer data and classification results are not updated regularly, customer credit states and market environment changes are not adapted, so that marketing schemes are not good in timeliness. These drawbacks restrict the accuracy, scientificity and sustainability of bank group marketing, and are difficult to meet the requirements of deep operation of banking customers, and improvement is needed. Disclosure of Invention The invention aims to provide a customer group accurate marketing evaluation classification method for data element analysis, which aims to solve the problems in the background technology. In order to achieve the above object, the present invention provides the following technical solutions: A customer group accurate marketing evaluation classification method for data element analysis comprises the following steps: S1, collecting credit report of banking client, and carrying out content identification and extraction on the credit report to construct client credit element information of banking client; s2, classifying the stock customers based on the credit factor information of the customers to obtain a plurality of customer group groups, wherein the customer group groups are associated with a single customer group label; S3, based on the credit element information of the customers, carrying out sensitivity classification on the banking customers by combining with the market contemporaneous reference interest rate to obtain the interest rate sensitivity level of each banking customer; s4, calculating the comprehensive score of the banking client based on the client credit element information, and dividing the marketing grade of the banking client according to the comprehensive score; S5, matching a basic differentiated marketing scheme through a rule engine based on customer group labels, marketing grades and interest rate sensitivity grades of the banking clients, obtaining optimal interest rate adjustment amplitude of the