A tobacco customer value evaluation system

Abstract

The present application relates to a kind of tobacco customer value evaluation system, belong to customer value evaluation technical field.The system includes data acquisition module and customer value evaluation module, data acquisition module is connected with customer value evaluation module;The data acquisition module is used to connect with tobacco industry integration platform, is used to obtain tobacco customer basic data from tobacco industry integration platform, and the text data serialization processing of tobacco customer basic data obtains tobacco customer value evaluation index data;The customer value evaluation module is used to obtain the comprehensive value of tobacco customer according to tobacco customer value evaluation index data, in combination with the tobacco customer value evaluation model constructed;The tobacco customer value evaluation model is based on tobacco customer value evaluation index score and tobacco customer value evaluation index weight to construct.The present application improves the efficiency of customer management and customer service quality.

Description

Technical Field

[0001] This invention belongs to the field of customer value evaluation technology, and specifically relates to a tobacco customer value evaluation system. Background Technology

[0002] Collecting and applying retail customer service needs is a crucial foundation for tobacco commercial enterprises to improve customer service satisfaction and strengthen channel control. By establishing a scientific and effective retail customer value classification management system and standard procedures for collecting service needs, precise classification services can be implemented for retail customers. This can effectively balance service implementation resources with personalized service needs, promote the transformation of customer service from "broad-based" to "precision-based," and drive high-quality development of industry services.

[0003] With the deepening of tobacco industry reforms, Customer Relationship Management (CRM) in tobacco companies is receiving increasing attention, with more and more activities and investments being made in this area. Therefore, how to improve the efficiency of customer management and the quality of customer service through Customer Value (CV) management is a pressing issue that tobacco companies urgently need to address and a common concern for many enterprises. In customer value management, the setting of customer value assessment indicators and the construction of evaluation models are key issues in the evaluation methodology. The existing practical tobacco customer value evaluation system remains an unresolved challenge.

[0004] Therefore, there is an urgent need for a tobacco customer evaluation system to solve the problems of low customer management efficiency and poor service quality faced by the tobacco industry. Summary of the Invention

[0005] The purpose of this invention is to provide a tobacco customer value evaluation system to solve the problems of low management efficiency and poor service quality in existing evaluation management systems.

[0006] To solve the above-mentioned technical problems, the present invention provides a tobacco customer value evaluation system, which includes a data acquisition module and a customer value evaluation module, wherein the data acquisition module is connected to the customer value evaluation module.

[0007] The data acquisition module is used to connect with the integrated tobacco industry platform to obtain basic tobacco customer data from the platform and to perform text data serialization processing on the basic tobacco customer data to obtain tobacco customer value evaluation index data. The tobacco customer value evaluation index includes a contribution index that characterizes the degree of contribution to the company's current revenue, a digital intelligence index that assesses the customer's responsiveness and acceptance of digital marketing, a credit index that examines the customer's credit status and repayment ability, a growth index that measures the customer's future development potential, and an influence index that assesses the customer's influence on the purchasing decisions of others.

[0008] The customer value evaluation module is used to obtain the comprehensive value of tobacco customers based on tobacco customer value evaluation index data and the constructed tobacco customer value evaluation model; the tobacco customer value evaluation model is constructed based on the tobacco customer value evaluation index scores and the weights of the tobacco customer value evaluation indexes.

[0009] Furthermore, the method for determining the weights of the tobacco customer value evaluation indicators is as follows: based on the tobacco customer value evaluation indicators, a five-level hierarchical structure model is constructed. According to the hierarchical structure model, a judgment matrix that reveals the importance and superiority / inferiority relationship between each element is constructed layer by layer. Through the analytic hierarchy process, the eigenvector corresponding to the largest eigenvalue of each judgment matrix is ​​determined, and finally the weight matrix of the tobacco customer value evaluation indicators is obtained.

[0010] Furthermore, a judgment matrix revealing the importance and superiority relationships among various elements is constructed layer by layer using the 1-9 scale method.

[0011] Furthermore, the constructed judgment matrix is ​​subjected to consistency verification to check whether the judgment matrix meets the preset consistency standard. If it does not meet the consistency standard, the judgment matrix is ​​modified until it meets the consistency standard.

[0012] Furthermore, when obtaining the weights of the tobacco customer value evaluation indicators, the scoring of the importance of the tobacco customer value evaluation indicators by tobacco industry experts is also taken into account.

[0013] Furthermore, the method for obtaining the scores of the tobacco customer value evaluation indicators is as follows: input the data of each tobacco customer value evaluation indicator into the corresponding scoring algorithm, and use a recurrent neural network to perform data calculation to obtain the score result of each indicator.

[0014] Furthermore, the system also includes a data application module, which is used to classify tobacco customers based on their overall value and pre-set classification criteria.

[0015] Furthermore, the classification criteria are divided into 5 levels according to their overall value, and each level has a corresponding marketing strategy and service plan.

[0016] Furthermore, the data application module also includes visualization processing functions, which are used to visualize the overall value of tobacco customers and / or customer classification results to generate intelligent BI data dashboards.

[0017] Furthermore, the data application module also includes a data push function, which is used to push the comprehensive value and / or classification results of tobacco customers to the service support backend.

[0018] The beneficial effects of the evaluation system of this invention are as follows: The tobacco customer value evaluation system of this invention can capture customer data from an integrated platform, combine it with the constructed customer value evaluation model to achieve a comprehensive evaluation of tobacco customer value, improve the efficiency of tobacco customer management through the management of tobacco customer value, provide data support for the implementation of precise classification services for retail customers, and thus effectively balance service implementation resources and personalized service needs, promote the transformation of customer service from "broad-based" to "precision-based", and promote the high-quality development of industry services. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the tobacco customer value hierarchy analysis model according to an embodiment of the present invention;

[0020] Figure 2 This is an example diagram illustrating the instructions for filling out the tobacco customer value evaluation form according to an embodiment of the present invention;

[0021] Figure 3 This is a schematic diagram illustrating the customer value weighting of a tobacco company according to an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the tobacco customer management platform architecture according to an embodiment of the present invention. Detailed Implementation

[0023] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0024] The core of this invention lies in: by setting tobacco customer value evaluation indicators and establishing a tobacco customer value evaluation model, a tobacco customer value evaluation system is constructed to improve the management efficiency and service quality of tobacco customers.

[0025] The tobacco customer value evaluation system of this invention includes a data acquisition module and a customer value evaluation module, which are connected to each other. The data acquisition module connects to an integrated tobacco industry platform to acquire basic tobacco customer data and performs text serialization processing on this data to obtain tobacco customer value evaluation index data. The customer value evaluation module uses the tobacco customer value evaluation index data, combined with a constructed tobacco customer value evaluation model, to obtain the comprehensive value of tobacco customers. The tobacco customer value evaluation model is constructed based on tobacco customer value evaluation index scores and weights.

[0026] Existing customer evaluation indicators and models mainly include the Net Present Value (NPV) customer evaluation system, the Recency Frequency (RFM) model, and the Current Value / Potential Value system. Current Value refers to realized customer value, i.e., the relationships and transactions already established between the company and its customers. It can be obtained by analyzing existing customer data, such as consumption behavior, purchase history, and customer satisfaction. Current Value helps companies understand their current market positioning and competitive advantages, and helps them better maintain existing customer relationships and improve customer satisfaction. Potential Value, on the other hand, refers to unrealized customer value, stemming from relationships and transactions yet to be established between the company and potential customers. Potential Value can be obtained by analyzing potential customer data, such as their demographic information, interests, and consumption habits. Potential Value can help companies explore new market opportunities and customer groups, expanding their market share and influence.

[0027] Research and analysis have shown that the current potential value system is highly applicable to the tobacco industry. By analyzing the consumption behavior and purchasing history of existing customers, tobacco companies can understand current market trends and hot topics, thereby adjusting product strategies and marketing. Simultaneously, by analyzing data on potential customers, tobacco companies can understand potential market opportunities, potential consumer habits and preferences, and thus explore new markets and customer groups. Furthermore, due to the involvement of health and hygiene issues in the tobacco industry, companies also need to consider social responsibility and ethical factors. Therefore, this embodiment, based on the current potential value system and combined with the diverse and personalized needs of tobacco customers, constructs a customer value evaluation system for the tobacco industry.

[0028] Through a detailed analysis of tobacco customer value evaluation indicators, these indicators are divided into two main factors: current value and potential value. Combining organizational and job analysis results, current value is further subdivided into three key indicator dimensions: contribution, digital intelligence, and creditworthiness, to more comprehensively assess the customer's existing value. Simultaneously, potential value is divided into two important indicator dimensions: growth potential and influence, to accurately predict the customer's future development potential. Current value includes contribution, digital intelligence, and creditworthiness. Contribution measures the customer's contribution to the company's current revenue, such as spending amount and purchase frequency. Digital intelligence assesses the customer's responsiveness and acceptance of digital marketing, such as online purchase rate and use of electronic payments. Creditworthiness examines the customer's credit status and repayment ability, such as credit rating and overdue payment history. Potential value includes growth and influence. Growth measures the customer's future growth potential, such as age and income growth rate. Influence assesses the customer's influence on others' purchasing decisions, such as word-of-mouth and social media influence. These indicator dimensions constitute a five-level hierarchical structure model, as follows: Figure 1 As shown.

[0029] Figure 1 This document presents a three-tiered value evaluation index for tobacco customers. The first tier comprises current value and potential value. Secondary indicators under current value include contribution, digital intelligence, and creditworthiness. Secondary indicators under potential value include growth and influence. Tertiary indicators under contribution include retailer gross profit contribution, order amount ratio, sales proportion of key brands, and low-tar order volume proportion. Tertiary indicators under digital intelligence include the configuration of digital management equipment, data integration mode, and data quality. Tertiary indicators under creditworthiness include compliance with laws and regulations and adherence to contracts. Tertiary indicators under growth include low-tar growth rate, single-pack value growth rate, sales growth rate of key brands, and sales growth rate of jointly developed products across the province. Tertiary indicators under influence include business district type, breadth of second-tier and above key brands, and terminal type. Table 1 provides a detailed explanation of the customer value evaluation indicators at each level.

[0030] Table 1: Explanation of Customer Value Evaluation Indicators

[0031]

[0032]

[0033] Based on the aforementioned three-level evaluation indicators, as a preferred implementation method, the data acquisition module can utilize distributed streaming data processing tools and Python to extract basic tobacco customer data from the integrated tobacco industry platform. As another preferred implementation method, the data acquisition module can use a pre-trained autoencoder to perform text data serialization processing on the extracted basic data to obtain all relevant data related to the aforementioned three-level indicators.

[0034] As a preferred implementation, the scoring algorithm for each indicator is embedded in the data analysis and processing module using transfer learning technology. After obtaining the relevant data for the required evaluation indicators, a recurrent neural network is used to process the data to obtain the scoring result for each indicator.

[0035] The tobacco customer value evaluation model in this embodiment is constructed based on tobacco customer value evaluation indicators and their weights. As a preferred implementation method, based on the advantages of comprehensively considering multiple factors, ensuring scientific reliability, high expert participation, and ease of operation and interpretation, this embodiment selects the Analytic Hierarchy Process (AHP) to calculate the weights of the tobacco customer value evaluation indicators. Specifically, based on the tobacco customer value evaluation indicators, a five-level hierarchical structure model is constructed. According to the hierarchical structure model, a judgment matrix revealing the importance and superiority relationships between various elements is constructed layer by layer. Through the AHP, the eigenvector corresponding to the largest eigenvalue of each judgment matrix is ​​determined, ultimately obtaining the weight matrix of the tobacco customer value evaluation indicators.

[0036] Judgment matrices provide a systematic and ordered method for more accurately analyzing and measuring the degree of correlation between objective things. Through judgment matrices, the basis for weight calculation can be obtained, thereby guiding the decision-making process, helping to make rational and reliable choices, and obtaining more accurate and reliable results in complex decision-making problems.

[0037] As a preferred implementation, a judgment matrix is ​​constructed based on the AHP method. Constructing the judgment matrix is ​​a crucial step in the AHP method, based on the design principles of hierarchical models, revealing the relative importance of each element through layer-by-layer comparisons. This process includes the following steps:

[0038] First, based on the hierarchical structure of the assessment, the various factors are organized according to their hierarchical relationships, forming a hierarchical structure diagram. The hierarchical structure diagram presents the decomposition and organization of the problem, showing the relative importance of the factors from top to bottom.

[0039] Secondly, based on the hierarchical structure diagram, a judgment matrix is ​​constructed layer by layer. Each element at each level is compared pairwise with the elements of the adjacent layer above it, using these elements as criteria.

[0040] The AHP method uses priority weights as an indicator to distinguish the superiority or inferiority of alternatives. Priority weights are relative metrics representing the relative merit of alternatives, with values ​​ranging from 0 to 1. Under a given decision criterion, a larger value indicates a superior alternative, and vice versa. The priority weights of each alternative at the alternative level with respect to the overall objective criterion system are calculated hierarchically from top to bottom. This process is called hierarchical weight analysis. In practice, the construction of the judgment matrix typically uses the 1-9 scale method referenced by Saaty.

[0041] In this embodiment, a 1-9 scale method is used to construct the judgment matrix. Two elements are compared and assigned relative weights using the 1-9 scale method, where 1 indicates equal importance, 3 indicates moderate importance, 5 indicates significant importance, 7 indicates strong importance, and 9 indicates extreme importance. The meanings of each scale level are shown in Table 2. Furthermore, if the relative importance of two elements is difficult to distinguish, intermediate values ​​of 2, 4, 6, and 8 can be used for fuzzy comparison.

[0042] Table 2: Scale and Meaning Description Table

[0043]

[0044]

[0045] For example, such as Figure 2 As shown, when comparing the low coke growth rate (A) and the single-value growth rate (B) pairwise, if A is equally important as B, fill in 1; if A is slightly more important than B, fill in 3; if A is relatively more important than B, fill in 5; if A is very important than B, fill in 7; if A is absolutely more important than B, fill in 9. The score range is 1-9 points. If B is equally important as A, fill in 1; if B is slightly more important than A, fill in 1 / 3; if B is relatively more important than A, fill in 1 / 5; if B is very important than A, fill in 1 / 7; if B is absolutely more important than A, fill in 1 / 9. The score range is 1-1 / 9 points.

[0046] Next, the judgment matrix is ​​solved. Using the analytic hierarchy process (AHP), the largest eigenvalue and corresponding eigenvector of each level of the judgment matrix are calculated. Combined with the assigned weights, the relative importance between elements in the upper and lower levels is determined. The eigenvector is the eigenvector corresponding to the largest eigenvalue of the judgment matrix, representing the weight ratio of each element. After obtaining the largest eigenvalue and its corresponding eigenvector, normalization is performed to obtain a hierarchically ordered weight vector, ultimately forming an indicator weight matrix for evaluating the current and potential value of a project.

[0047] The first-level indicator weights represent the relative importance of each primary indicator in the evaluation. These primary indicators include contribution, impact, credibility, growth, and digital intelligence, etc., and they have different weights when evaluating the value of a project. By comprehensively considering their weights, the final evaluation result can be obtained.

[0048] The second-level indicator weights represent the relative importance of each secondary indicator in the evaluation. These secondary indicators include retailer contribution to gross profit, order amount share, key brand sales share, low-tar order volume share, configuration of digital management equipment, data integration mode, data quality, compliance with laws and regulations, contract compliance, low-tar growth rate, single-item value growth rate, key brand sales growth rate, province-wide sales growth rate of jointly cultivated product specifications, business district type, breadth of second-tier and above key brands, and terminal type, etc. They have different correlations and weights with the primary indicators. By comprehensively considering their weights, the weights of each primary indicator can be derived.

[0049] By evaluating the weights of the indicators, a quantitative analysis and comparison of the current and potential value of a project can be conducted. The higher the weight of each indicator, the greater its impact on the project; conversely, the lower the weight, the smaller its impact.

[0050] Taking the tobacco customer evaluation indicators of a certain tobacco company as an example, Figure 3 This section presents the various alternative options and their weights in relation to the decision objective. Option-level elements refer to the components of different plans proposed to achieve a specific decision objective. These option-level elements may have varying degrees of influence on the decision objective; ranking and assigning weights to them helps decision-makers better understand their relative importance.

[0051] Table 3 shows the weighting table for evaluating the tobacco customer value of this tobacco company. Each alternative option in the table has a weight value, representing the ranking weight of that option-level element to the decision objective. A larger weight value indicates a greater impact of that option-level element on the decision objective. Retailer contribution gross profit is the most important option-level element, with a weight value of 0.2833. This means that retailer contribution gross profit has the most significant impact on the decision objective. Compliance with laws and regulations is the second most important option-level element, with a weight value of 0.2208. Compliance with laws and regulations also has a very important impact on the decision objective. The weight values ​​of option-level elements such as the proportion of sales revenue from key brands and the proportion of order amount are relatively small, at 0.0977 and 0.0856 respectively. This indicates that their impact on the decision objective is relatively small. The weight values ​​of option-level elements such as data quality, compliance rate, and single-value growth rate are even lower, showing that their impact on the decision objective is smaller.

[0052] Table 3: Ranking Weights of Elements in the Decision-Making Objective at the Alternative Layer

[0053]

[0054]

[0055] As a preferred implementation, a consistency test is used to evaluate the reliability of the judgment matrix. The consistency test ensures that expert judgments remain consistent throughout the matrix. If the consistency ratio (CR) of the judgment matrix is ​​less than a predefined threshold, the matrix is ​​considered reliable, the evaluation results have high credibility, and the consistency requirement is met; the threshold is typically set to 0.1. If the consistency ratio (CR) of the judgment matrix is ​​greater than the predefined threshold, the matrix does not meet the consistency requirement, and the judgment matrix is ​​revised until the consistency standard is met. This analytical method ensures that trade-offs and choices are based on reliable data during the decision-making process, improving the accuracy and stability of the decision results.

[0056] As another implementation method, the weight of each evaluation indicator can also be determined through expert opinion surveys, data analysis, or experience summaries.

[0057] After obtaining the weights of customer value evaluation indicators based on the Analytic Hierarchy Process (AHP), each indicator is assigned a maximum score of 100 points. Using a scoring algorithm, the score result of each indicator is obtained. Therefore, the comprehensive value of a tobacco customer = the score result of each indicator * the corresponding weight.

[0058] As a preferred implementation, the tobacco customer value evaluation system in this embodiment also includes a data application module. This module categorizes customers based on their overall value and pre-defined classification criteria. Customer segmentation helps companies manage customers in a categorized manner, thereby maintaining or expanding their existing customer base and increasing their value. Furthermore, it enhances the company's competitiveness and resource allocation efficiency, improving operational effectiveness. Additionally, it allows for the provision of different resources and services to different customer groups. This differentiated service enhances customer loyalty and strengthens the bond between the company and its core customers, enabling them to contribute more value to the company.

[0059] In a preferred implementation, this embodiment divides customers into 5 levels, each with corresponding marketing strategies and service plans, as shown in Table 4. 0-20%: This category of customers scores the lowest, indicating a low ranking in customer value assessment. 20%-40%: This category of customers scores low, but slightly higher than the first category. 40%-60%: This category of customers scores at a moderate level, showing a certain level of satisfaction and loyalty to the company's products or services. Their purchase frequency and spending are also relatively high. These customers are stable consumers for the company, and maintaining and serving them is very important. 60%-80%: This category of customers scores high, indicating a high level in customer value assessment. 80%-100%: This category of customers scores the highest, representing the highest value level in customer value assessment. They are the company's core customers, with very high purchase frequency and spending, and exhibit extremely high loyalty and satisfaction.

[0060] Table 4: Five-Tier Customer Classification

[0061] Star-rated service Service Menu Customer types One-star service Level 1 service quality and efficiency, standardized service content 0-20％ Two-star service Level 2 service quality and efficiency, standardized service content 20％-40％ Samsung Services Level 2 service quality and efficiency: standardized services plus some personalized services. 40％-60％ Four-star service Level 3 service quality and efficiency, standardized services plus some personalized services. 60％-80％ Five-star service Level 3 service quality and efficiency, standardized services plus all personalized service content. 80％-100％

[0062] Customer segmentation results are mainly used in business guidance services, classified supply of cigarettes, allocation of cigarette resources and brand cultivation. By segmenting customers, more targeted and effective marketing strategies and services can be provided to retail customers, thereby improving customer satisfaction.

[0063] In a preferred implementation, the data application module also includes visualization processing capabilities to visualize and display the overall value and / or customer classification results of tobacco customers. Preferably, the evaluation results are visualized using tools such as SPSS and Worktile to generate an intelligent BI (Business Intelligence) data dashboard.

[0064] As a preferred implementation, the data application module also includes a data push function, which is used to push the comprehensive value and / or classification results of tobacco customers to the service support backend, so that the service support backend can formulate corresponding business guidance strategies based on the customer classification results of the data dashboard, and then push the generated digital business guidance document to the service execution frontend through the service middle platform, so that the service personnel at the service execution frontend can carry out customer guidance.

[0065] Based on the aforementioned tobacco customer value evaluation system, this invention also provides a tobacco customer management platform. For example... Figure 4As shown, the management platform includes a service execution front-end, a service middle platform, and a service support back-end connected sequentially. The service front-end interacts with customer terminals, provides service guidance, and collects tobacco customer data. The service middle platform includes an integrated tobacco industry platform and the tobacco customer value evaluation system. The integrated tobacco industry platform is connected to both the service execution front-end and the customer value evaluation module in the tobacco customer value evaluation system, and is used to obtain and store tobacco customer data from the service execution front-end. The service support back-end is used to formulate corresponding business guidance strategies based on the customer value evaluation data analyzed by the customer value evaluation module in the tobacco customer value evaluation system. The tobacco customer value evaluation system uses the evaluation system described in the above embodiment, which will not be described in detail here.

[0066] This management platform is a four-in-one service platform system built on platform ecosystem theory and deep learning technology, applying platform construction technologies such as development technologies (Java, Python), database technologies (MySQL), and cloud computing technologies (Azure). The system centers on a service middle platform, with service specialists and terminal specialists forming the service execution front end, and brand specialists and digital specialists forming the service support back end, thus forming a closely collaborative platform ecosystem.

[0067] The integrated platform for the tobacco industry comprises subsystems such as a POS terminal store management system and a mobile office system for marketing personnel. Based on blockchain technology, it enables the capture and aggregation of data from multiple endpoints within these subsystems. As a distributed ledger, blockchain ensures data security and traceability. This allows decision-makers and executives within tobacco companies to share and verify data through the blockchain network, thereby establishing trust and transparency. Furthermore, the system utilizes smart contracts to store data on the blockchain and automatically trigger corresponding operations. For example, when the POS terminal store management system generates new transaction data, the smart contract can automatically update the relevant blockchain ledger and push the data to other related platform subsystems. This distributed streaming data application ensures the real-time nature and accuracy of the data.

[0068] When the service platform uses tools such as SAS to generate visualized intelligent data BI dashboards, the service support backend can perform data analysis and judgment based on the data on the blockchain. Through the immutability and data integrity provided by blockchain, the service support backend can conduct data analysis more reliably, improving the accuracy and efficiency of decision-making.

[0069] The service support backend pushes formulated operational guidance strategies to the service execution frontend through the service middle platform, enabling service personnel at the frontend to provide customer guidance. Instructions are transmitted from the service middle platform to the service execution frontend, forming a four-in-one PDCA closed-loop linkage model for comprehensive service management and optimization. The service middle platform, acting as a hub, transmits data and instructions to the service execution frontend via blockchain technology. Through deep learning technology, it can analyze large amounts of data, allowing the service execution frontend to make intelligent decisions and personalized service recommendations based on the instructions from the service middle platform. By identifying and understanding user needs, this management platform can provide accurate solutions. Combined with the data sharing and smart contract execution capabilities of blockchain, it can achieve efficient service delivery and optimization, improve service quality and efficiency, and bring a better service experience to retail customers.

[0070] Therefore, based on the customer value evaluation model, the management platform in this embodiment fully utilizes digital platforms and tools to establish a service platform ecosystem consisting of a service execution front-end, a service middle platform, and a service support back-end. By collecting data at the service execution front-end, processing data at the service middle platform, and analyzing data at the service support back-end before pushing decisions back to the front-end, a PDCA management closed loop is formed. This improves the efficiency of customer management and enables precise classification services for retail customers. It effectively balances service implementation resources with personalized service needs, promotes the transformation of customer service from "broad-based" to "precision-based," and drives high-quality development of industry services.

Claims

1. A tobacco customer value evaluation system, characterized in that, The system includes a data acquisition module and a customer value evaluation module, with the data acquisition module connected to the customer value evaluation module; The data acquisition module is used to connect with the integrated tobacco industry platform to obtain basic tobacco customer data from the platform and to perform text data serialization processing on the basic tobacco customer data to obtain tobacco customer value evaluation index data. The tobacco customer value evaluation index includes a contribution index that characterizes the degree of contribution to the company's current revenue, a digital intelligence index that assesses the customer's responsiveness and acceptance of digital marketing, a credit index that examines the customer's credit status and repayment ability, a growth index that measures the customer's future development potential, and an influence index that assesses the customer's influence on the purchasing decisions of others. The customer value evaluation module is used to obtain the comprehensive value of tobacco customers based on tobacco customer value evaluation index data and the constructed tobacco customer value evaluation model; the tobacco customer value evaluation model is constructed based on the tobacco customer value evaluation index scores and the weights of the tobacco customer value evaluation indexes.

2. The tobacco customer value evaluation system according to claim 1, characterized in that, The method for determining the weights of the tobacco customer value evaluation indicators is as follows: Based on the tobacco customer value evaluation indicators, a five-level hierarchical structure model is constructed. According to the hierarchical structure model, a judgment matrix that reveals the importance and superiority relationship between each element is constructed layer by layer. Through the analytic hierarchy process, the eigenvector corresponding to the largest eigenvalue of each judgment matrix is ​​determined, and finally the weight matrix of the tobacco customer value evaluation indicators is obtained.

3. The tobacco customer value evaluation system according to claim 2, characterized in that, A judgment matrix revealing the importance and superiority relationships among various elements is constructed layer by layer using the 1-9 scale method.

4. The tobacco customer value evaluation system according to claim 2, characterized in that, The constructed judgment matrix is ​​subjected to consistency verification to check whether it meets the preset consistency standard. If it does not meet the consistency standard, the judgment matrix is ​​modified until it meets the consistency standard.

5. The tobacco customer value evaluation system according to claim 2, characterized in that, When determining the weights of the tobacco customer value evaluation indicators, the scoring of the importance of these indicators by tobacco industry experts was also taken into account.

6. The tobacco customer value evaluation system according to claim 1, characterized in that, The method for obtaining the scores of the tobacco customer value evaluation indicators is as follows: input the data of each tobacco customer value evaluation indicator into the corresponding scoring algorithm, use a recurrent neural network to perform data calculation, and obtain the score result of each indicator.

7. The tobacco customer value evaluation system according to claim 1, characterized in that, The system also includes a data application module, which is used to classify tobacco customers based on their overall value and pre-set classification criteria.

8. The tobacco customer value evaluation system according to claim 7, characterized in that, The classification criteria are divided into 5 levels according to their overall value, and each level has a corresponding marketing strategy and service plan.

9. The tobacco customer value evaluation system according to claim 7, characterized in that, The data application module also includes visualization processing functions, which are used to visualize the overall value of tobacco customers and / or customer classification results to generate intelligent BI data dashboards.

10. The tobacco customer value evaluation system according to claim 7, characterized in that, The data application module also includes a data push function, which is used to push the comprehensive value and / or classification results of tobacco customers to the service support backend.

Images