A customer demand management method and system based on big data analysis

Abstract

The application relates to the technical field of customer demand management, and particularly discloses a customer demand management method and system based on big data analysis, which comprises the following steps: S1, collecting image data of a customer through a collection module; S2, analyzing and extracting facial features and dressing images of the customer through a feature extraction module, and preliminarily predicting the character and demand of the customer based on big data analysis technology through a big data analysis module, and pushing the preliminary prediction result to a service personnel; and S3, in a service process, real-time analyzing and extracting facial expression and body language features of the customer through the feature extraction module, and real-time analyzing through an emotion analysis module, real-time monitoring the emotional state of the customer, and assisting the service personnel in service; the application collects image data of a customer, preliminarily predicts the character and demand of the customer based on big data analysis, and pushes the preliminary prediction result to a service personnel, so that the service personnel can better grasp the demand of the customer.

Description

Technical Field

[0001] This invention relates to the field of customer demand management technology, specifically to a customer demand management method and system based on big data analysis. Background Technology

[0002] With the booming development of the online economy, traditional retail has suffered an unprecedented impact. In order to stand firm in the fierce market competition, offline stores have gradually shifted their core competitiveness to personalized service capabilities. However, there are still many shortcomings in the current management of customer needs of offline stores.

[0003] Traditional customer needs management relies heavily on the experience and judgment of service personnel, which is highly subjective and lacks precision. It struggles to quickly and accurately grasp customers' true needs and emotional changes. For example, relying solely on visual observation makes it difficult to capture subtle emotional signals conveyed by customers' facial expressions and body language, let alone conduct in-depth analysis based on large amounts of data. Furthermore, service strategies are often simplistic and unable to be flexibly adjusted according to the customer's real-time emotional state, leading to poor customer experience and customer churn. Simultaneously, the lack of a scientific and reasonable mechanism for the allocation and coordination of service personnel hinders the full utilization of team strengths and the improvement of overall service quality. Summary of the Invention

[0004] The purpose of this invention is to provide a customer demand management method and system based on big data analysis, and to solve the following technical problems:

[0005] The question is how to leverage advanced technology to accurately analyze customer needs and emotions in order to enhance the personalized service capabilities of offline stores.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A customer demand management method based on big data analytics, the method comprising:

[0008] S1. Collect customer image data through the acquisition module;

[0009] S2. The feature extraction module analyzes and extracts the customer's facial features and clothing images. The big data analysis module uses big data analysis technology to make a preliminary prediction of the customer's personality and needs, and pushes the preliminary prediction results to the service personnel.

[0010] S3. During the service process, the feature extraction module analyzes and extracts the customer's facial expressions and body language features in real time, and analyzes them in real time through the emotion analysis module to monitor the customer's emotional state in real time and assist service personnel in providing services.

[0011] S4. The early warning analysis module continuously monitors the customer's cumulative emotional value within the sliding time window and compares the cumulative emotional value with the preset emotional threshold range. When the cumulative emotional value is greater than the maximum value of the preset emotional threshold range, the emotional analysis module continuously assists the service personnel in providing services. When the cumulative emotional value is within the preset emotional threshold range, the emotional analysis module switches to the manual intervention mode, and other service personnel in the store intervene to provide collaborative services. When the cumulative emotional value is less than the minimum value of the preset emotional threshold range, a service handover instruction is automatically generated to prompt other service personnel in the store to take over the service work for the current customer.

[0012] Furthermore, in step S2, the process by which the big data analysis module makes a preliminary prediction of customer personality based on big data analysis technology includes:

[0013] Extract facial features from the customer's facial image and construct a facial feature set Q;

[0014] The set of facial features Q is matched with a rule library of facial features-personality associations built based on big data analysis;

[0015] If the similarity between the feature combination in Q and a certain set of facial feature combinations in the rule base reaches a preset similarity threshold, then the personality type corresponding to that set of facial feature combinations will be used as the preliminary prediction result of the customer's personality.

[0016] If the similarity between the feature combination in Q and any set of facial feature combinations in the rule base does not reach the preset similarity threshold, then the personality types corresponding to the top three sets of facial feature combinations in the rule base with the highest similarity to the feature combination in Q will be used as the preliminary prediction results of the customer's personality.

[0017] Furthermore, in step S2, the process by which the big data analysis module makes a preliminary prediction of customer demand based on big data analysis technology includes:

[0018] Segment customer clothing images, identify clothing style, color, and material characteristics, and construct a clothing feature set W;

[0019] Calculate the similarity between the clothing feature set W and each combination of clothing features in the clothing-demand association rule base established based on big data analysis;

[0020] The product categories corresponding to the most similar clothing feature combinations are used as a preliminary prediction of customer demand.

[0021] Furthermore, in step S3, the process by which the emotion analysis module monitors the customer's emotional state in real time includes:

[0022] Identify and extract customer facial expression and body language features;

[0023] Through formula The emotional characteristic value R of the customer is obtained through analysis and calculation;

[0024] Where N is the number of preset facial expression features. , Let M be the emotion rating value of the i-th facial expression feature, and M be the preset number of body language features. , Let j be the emotion rating value of the j-th body language feature. , This is the first weighting coefficient;

[0025] A Cartesian coordinate system is established with timestamps as the X-axis and the obtained customer emotional feature values ​​R as the Y-axis. A smooth curve is then used to sequentially connect the distribution points formed by each time point and emotional feature value during the customer service process, thus forming a curve representing the customer's emotional state. .

[0026] Furthermore, in step S4, the process by which the early warning analysis module continuously monitors the customer's cumulative sentiment value within the sliding time window includes:

[0027]

[0028] The cumulative emotional value of customers within the sliding time window is obtained by analyzing and calculating using formulas (1)-(2). ;

[0029] in, For sliding time windows, The curve showing how a client's negative emotional value changes over time. , This is the second weighting coefficient;

[0030] Accumulated emotional value Compared with the preset emotional threshold range Perform a comparison;

[0031] when At the same time, the sentiment analysis module continuously assists service personnel in providing services;

[0032] when At this time, other service personnel in the store will intervene to provide assistance;

[0033] when When this happens, other staff members in the store should be notified to take over the service for the current customer.

[0034] Furthermore, the maximum value of the preset emotion threshold range can be intelligently adjusted in the following ways:

[0035] Establish a service personnel evaluation index system to quantitatively score service personnel through three dimensions: service duration, service quality, and on-site performance.

[0036] Through formula The analysis and calculation yielded the maximum value within the preset emotional threshold range. ;

[0037] in, This is the base value for the preset maximum range of emotion thresholds. For adjustment coefficients, X represents the cumulative service time of the service personnel, X represents the historical service quality evaluation value of the service personnel, and C represents the on-site performance of the service personnel during the current service process. , , This is the third weighting coefficient.

[0038] Furthermore, the screening process for other service personnel includes:

[0039] Establish a service personnel screening index system, and quantitatively score service personnel through three dimensions: service experience, service evaluation, and service authority;

[0040] Through formula The analysis and calculation yielded a comprehensive score for each service personnel. ;

[0041] in, The service experience evaluation score for service personnel. The service evaluation score for service personnel The service authority evaluation value for service personnel. , , This is the fourth weighting coefficient;

[0042] Service personnel are ranked from highest to lowest based on their overall scores, and those with higher overall scores are selected to provide collaborative services or take over the current customer's service work; when overall scores are the same, service personnel with higher service authority levels are given priority.

[0043] A customer demand management system based on big data analytics, wherein the system is applied to a customer demand management method based on big data analytics, and the system includes:

[0044] The data acquisition module is used to collect image data from customers and service-related data from service personnel.

[0045] The feature extraction module is used to analyze and extract customer facial features, clothing images, facial expressions, and body language features;

[0046] The big data analytics module is used to perform big data analysis based on customers' facial features and clothing images to make preliminary predictions about customers' personality and needs, in order to assist service personnel in providing services.

[0047] The emotion analysis module is used to monitor the emotional state of customers in real time based on their facial expressions and body language characteristics, in order to assist service personnel in providing services.

[0048] The early warning analysis module is used to continuously monitor the cumulative emotional value of customers within a sliding time window, compare it with a preset emotional threshold range, and trigger different service response mechanisms based on the comparison results; intelligently adjust the maximum value of the preset emotional threshold range; and screen other service personnel when selecting other service personnel to intervene in collaborative services or take over the service work of the current customer.

[0049] The communication module is used to transmit information such as the preliminary prediction results of customer personality and needs from the big data analysis module, the monitoring results of customer emotional state from the emotion analysis module, and the service response instructions from the early warning analysis module to the headset worn by the service personnel in real time.

[0050] The beneficial effects of this invention are:

[0051] (1) This invention collects customer image data, makes preliminary predictions on customer personality and needs based on big data analysis, and pushes the preliminary prediction results to service personnel to improve service personnel’s grasp of customer needs. During the service process, the emotional state of the customer is monitored in real time and the emotional state is pushed to service personnel to help service personnel provide better service. Different service responses are triggered based on the customer’s accumulated emotional value, which can accurately grasp customer needs and thus improve service quality and customer satisfaction.

[0052] (2) By establishing a service personnel evaluation index system, this invention comprehensively considers factors such as service duration, service quality and on-site performance of service personnel, and dynamically adjusts the maximum value of the preset emotional threshold range, so that the service response mechanism is more in line with the actual ability of service personnel and improves the scientificity and flexibility of service management.

[0053] (3) By establishing a service personnel screening index system, this invention comprehensively considers factors such as service experience, service evaluation and service authority of service personnel, calculates comprehensive scores and sorts and screens them, and can select the most suitable personnel when other service personnel need to intervene, so as to ensure the continuity and high quality of service. Attached Figure Description

[0054] The invention will now be further described with reference to the accompanying drawings.

[0055] Figure 1This is a flowchart illustrating the steps of a customer demand management method based on big data analysis proposed in this invention.

[0056] Figure 2 This is a schematic diagram of a customer demand management system based on big data analysis proposed in this invention. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Please see Figure 1 As shown, in one embodiment, a customer demand management method based on big data analytics is provided, the method comprising:

[0059] S1. Collect customer image data through the acquisition module. The acquisition module can be an image acquisition device such as a high-definition camera installed in the store. The camera captures images of customers entering the store and obtains image data containing information such as the customer's face, clothing, expression and body language.

[0060] S2. The feature extraction module analyzes and extracts the customer's facial features and clothing images. The big data analysis module uses big data analysis technology to make a preliminary prediction of the customer's personality and needs, and pushes the preliminary prediction results to the service personnel.

[0061] S3. During the service process, the feature extraction module analyzes and extracts the customer's facial expressions and body language features in real time, and analyzes them in real time through the emotion analysis module to monitor the customer's emotional state in real time and assist service personnel in providing services.

[0062] S4. The early warning analysis module continuously monitors the customer's cumulative emotional value within the sliding time window and compares the cumulative emotional value with the preset emotional threshold range. When the cumulative emotional value is greater than the maximum value of the preset emotional threshold range, the emotional analysis module continuously assists the service personnel in providing services. When the cumulative emotional value is within the preset emotional threshold range, the emotional analysis module switches to the manual intervention mode, and other service personnel in the store intervene to provide collaborative services. When the cumulative emotional value is less than the minimum value of the preset emotional threshold range, a service handover instruction is automatically generated to prompt other service personnel in the store to take over the service work for the current customer.

[0063] Through the above technical solution, this embodiment provides a customer demand management method based on big data analysis. The method collects customer image data, makes preliminary predictions about customer personality and needs based on big data analysis, and pushes the preliminary prediction results to service personnel to improve service personnel's understanding of customer needs. During the service process, the method monitors the customer's emotional state in real time and pushes the emotional state to service personnel to assist them in providing better service. Different service responses are triggered based on the customer's accumulated emotional value, which can accurately grasp customer needs and thus improve service quality and customer satisfaction.

[0064] In one embodiment, step S2, the process by which the big data analysis module makes a preliminary prediction of customer personality based on big data analysis technology, includes:

[0065] Extract facial features from customer facial images and construct facial feature set Q. The extraction of facial features from customer facial images can utilize image processing and computer vision technologies, such as using key point detection algorithms to identify the contour, shape, size, and other features of customer facial features.

[0066] The set of facial features Q is matched with the facial feature-personality association rule library established based on big data analysis. The facial feature-personality association rule library can be obtained by collecting facial feature image data of many people with known personality types and using algorithms such as cluster analysis and machine learning to mine the association between facial feature combinations and personality types, or it can be established manually based on experience.

[0067] If the similarity between the feature combination in Q and a certain set of facial feature combinations in the rule base reaches a preset similarity threshold, the preset similarity threshold can be adjusted according to the actual situation and experimental results, such as setting it to 80%, then the personality type corresponding to that set of facial feature combinations is used as the preliminary prediction result of the customer's personality.

[0068] If the similarity between the feature combination in Q and any set of facial feature combinations in the rule base does not reach the preset similarity threshold, then the personality types corresponding to the top three sets of facial feature combinations in the rule base with the highest similarity to the feature combination in Q will be used as the preliminary prediction results of the customer's personality.

[0069] In step S2, the process by which the big data analysis module makes a preliminary prediction of customer demand based on big data analysis technology includes:

[0070] Segmenting customer clothing images can be achieved using image segmentation algorithms to separate customer clothing from the overall image and identify clothing style, color, and material features. The identification of clothing style, color, and material features can be aided by deep learning models, such as convolutional neural networks, for training and recognition, to construct a clothing feature set W.

[0071] Calculate the similarity between the clothing feature set W and each combination of clothing feature in the clothing-demand association rule base built based on big data analysis. The clothing-demand association rule base is also built based on a large amount of customer purchase data, clothing data, etc., through big data analysis. It collects customer purchase records and their clothing information at the time, and analyzes the relationship between clothing features and purchased product categories through data mining algorithms to form the rule base; similarity can be calculated using algorithms such as cosine similarity.

[0072] The product categories corresponding to the most similar clothing feature combinations are used as a preliminary prediction of customer demand.

[0073] Through the above technical solution, this embodiment provides a method for preliminary prediction of customer personality and needs based on big data analysis. The method extracts the customer's facial features and clothing features, constructs a set of facial features Q and a set of clothing features W, respectively, and performs matching and similarity calculation with a rule library of facial features-personality association and a rule library of clothing-needs association based on big data. It can preliminarily judge the customer's personality type and potential needs based on the customer's external characteristics, providing important reference for service personnel to formulate targeted service strategies and prepare product information in advance, effectively improving the adaptability and accuracy of services, and enhancing the customer service experience.

[0074] In one embodiment, step S3, the process of the emotion analysis module monitoring the customer's emotional state in real time includes:

[0075] Identify and extract customer facial expression and body language features;

[0076] Through formula The emotional characteristic value R of the customer is obtained through analysis and calculation;

[0077] Where N is the preset number of facial expression features, and a suitable value can be determined based on actual research and experiments, for example, set to 10 common facial expression features. , Let M be the emotional evaluation value of the i-th facial expression feature. Various facial expression features can be manually assigned corresponding emotional evaluation scores. Facial expression features reflecting positive emotions correspond to positive emotional evaluation values, while those reflecting negative emotions correspond to negative emotional evaluation values. For example, a frowning expression can be set to an emotional evaluation value of -0.8, and a smiling expression can be set to an emotional evaluation value of 0.7. M is the preset number of body language features, which can be set according to actual needs, such as 8. , Let be the emotional evaluation value of the j-th body language feature. Various body language expressions can be manually assigned corresponding emotional evaluation scores. This will reflect positive emotions through positive body language and negative emotions through negative body language. For example, a nodding body language can be set to an emotional evaluation value of 0.5, and a shaking body language can be set to an emotional evaluation value of -0.4. , This is the first weighting coefficient, which can be set based on experience;

[0078] A Cartesian coordinate system is established with timestamps as the X-axis and the obtained customer emotional feature values ​​R as the Y-axis. A smooth curve is then used to sequentially connect the distribution points formed by each time point and emotional feature value during the customer service process, thus forming a curve representing the customer's emotional state. When the customer's emotional characteristic value R is above the X-axis, it indicates that the customer's emotional state is positive; when the customer's emotional characteristic value R is below the X-axis, it indicates that the customer's emotional state is negative.

[0079] Through the above technical solution, this embodiment provides a method for real-time monitoring of customer emotional state. The method identifies customer facial expressions and body language features, calculates emotional feature values ​​using a specific formula, and plots emotional state curves. This allows for an intuitive and real-time presentation of customer emotional changes during the service process, assisting service personnel in adjusting service methods in a timely manner.

[0080] In one embodiment, step S4, where the early warning analysis module continuously monitors the customer's cumulative sentiment value within a sliding time window, includes:

[0081]

[0082]

[0083] The cumulative emotional value of customers within the sliding time window is obtained by analyzing and calculating using formulas (1)-(2). ;

[0084] in, This is a sliding time window, the size of which can be set according to the actual service scenario and needs, for example, it can be set to 10 minutes, 15 minutes, etc. This is a curve showing how a customer's negative emotional value changes over time, based on the customer's emotional characteristic values. The analysis and processing yielded the following results: , This is the second weighting coefficient, which can be optimized through experiments and data analysis;

[0085] Accumulated emotional value Compared with the preset emotional threshold range By comparison, the preset emotion threshold range can be obtained by preset based on experience;

[0086] when At the same time, the sentiment analysis module continuously assists service personnel in providing services;

[0087] when If the current customer is not very satisfied with the service provided by the current service personnel, other service personnel in the store will intervene to provide collaborative services. The intervention of other service personnel can be divided into several types, such as other service personnel making suggestions to the current service personnel in the background through wireless communication devices, or other service personnel appearing on-site to cooperate with the current service personnel. The specific choice depends on the settings of the store management personnel.

[0088] when This indicates that the current customer is likely dissatisfied with the service provided by the current staff member. If the current staff member continues to provide service, they may lose the current customer. This prompts another staff member in the store to take over the service for the current customer.

[0089] Through the above technical solution, this embodiment provides a customer service early warning method based on accumulated emotion value. The method monitors the customer's accumulated emotion value within a sliding time window and compares it with a preset threshold range. It triggers corresponding service responses according to different situations, which can respond to changes in customer emotions in a timely manner, thereby accurately analyzing customer needs, ensuring service quality, and improving customer experience.

[0090] In one embodiment, the maximum value of the preset emotion threshold range can be intelligently adjusted in the following way:

[0091] Establish a service personnel evaluation index system, which quantifies and scores service personnel through three dimensions: service duration, service quality, and on-site service performance. Service duration can be obtained by recording the cumulative time that service personnel have served customers through the system; historical service quality evaluation values ​​can be calculated based on customer ratings of past services, combined with factors such as the number of ratings; and on-site service performance can be obtained through on-site management personnel's observation and scoring, as well as real-time customer feedback (such as the number of times customer needs were misjudged).

[0092] Through formula The analysis and calculation yielded the maximum value within the preset emotional threshold range. ;

[0093] in, This serves as the baseline value for the preset maximum range of emotional thresholds. An initial value can be set based on industry norms or the company's historical experience. To adjust the adjustment coefficient, an optimization model incorporating multiple objectives such as customer satisfaction and service efficiency is established. This model is then used for regression analysis or machine learning training based on historical service data to determine the appropriate adjustment coefficient value. Alternatively, A / B testing can be conducted to compare service performance under different coefficients and identify the optimal value. X represents the cumulative service time of service personnel, i.e., the total time service personnel have participated in customer service work since joining the company. X represents the historical evaluation value of service quality of service personnel. Specifically, it can be calculated by integrating data such as customer proactive evaluations (e.g., satisfaction scores, written feedback), quality inspection scores (evaluated by the internal quality inspection team based on service audio / video recordings), and complaint records, according to certain scoring rules (e.g., weighted average). C represents the service performance level of service personnel in the current service process. Specifically, it can be identified through real-time voice analysis technology, behavior monitoring systems, etc., identifying on-site performance indicators such as tone of voice, speech rate, use of professional terminology, and problem-solving speed of service personnel, and converting them into a performance level value through a preset algorithm model. , , As the third weighting coefficient, multiple preset emotion threshold calculation schemes with different weight combinations can be designed and piloted on a small scale in actual business scenarios. By comparing key indicators such as customer complaint rate and service dispute handling efficiency under each scheme, statistical methods can be used to determine the optimal weighting coefficients for different evaluation dimensions in the calculation.

[0094] Through the above technical solution, this embodiment provides an intelligent adjustment method for the maximum value of a preset emotional threshold range. The method establishes a service personnel evaluation index system, comprehensively considers factors such as service duration, service quality, and on-site performance of service personnel, and dynamically adjusts the maximum value of the preset emotional threshold range, so that the service response mechanism is more in line with the actual ability of service personnel and improves the scientificity and flexibility of service management.

[0095] In one embodiment, the screening process for other service personnel includes:

[0096] Establish a service personnel screening index system, and quantitatively score service personnel through three dimensions: service experience, service evaluation, and service authority. Among them, the service experience evaluation value can be quantified based on factors such as the number of years the service personnel have engaged in relevant service work and the number of projects they have participated in; the service evaluation value can be calculated based on the historical evaluation scores of customers for the service personnel; and the service authority evaluation value can be set with different levels of scores based on the service personnel's position and authority within the company.

[0097] Through formula The analysis and calculation yielded a comprehensive score for each service personnel. ;

[0098] in, The service experience evaluation score for service personnel can be obtained through the following methods: statistically analyzing data such as the number of years the service personnel have engaged in relevant services, the number of projects they have participated in, and the number of times they have handled complex problems. This data is then weighted and calculated according to a pre-defined scoring model. The service evaluation score for service personnel is specifically based on customer feedback on the service process, including service attitude, response speed, and problem-solving effectiveness. Customer ratings can be collected through online evaluation systems, questionnaires, etc., and then aggregated and weighted. The service authority evaluation score for service personnel can be determined by assessing factors such as their job authority, resource allocation capabilities, and decision-making influence within the company. This score is assigned by company management or relevant departments based on established evaluation criteria. , , The fourth weighting coefficient is used to determine the relative importance of service experience, service evaluation, and service authority in the calculation of the overall score. It can be obtained and adjusted through actual business needs and data analysis.

[0099] Service personnel are ranked from highest to lowest based on their overall scores, and those with higher overall scores are selected to provide collaborative services or take over the current customer's service work; when overall scores are the same, service personnel with higher service authority levels are given priority.

[0100] Through the above technical solution, this embodiment provides a method for screening other service personnel. The method establishes a service personnel screening index system, comprehensively considers factors such as service experience, service evaluation and service authority of service personnel, calculates a comprehensive score and sorts and screens them, and can select the most suitable personnel when other service personnel need to intervene, so as to ensure the continuity and high quality of service.

[0101] Please see Figure 2 As shown, in one embodiment, a customer demand management system based on big data analytics is provided. The system is applied to a customer demand management method based on big data analytics, and includes:

[0102] The data acquisition module is used to collect image data of customers and service-related data of service personnel. The hardware of the data acquisition module can be multiple high-definition cameras distributed in different areas of the store to collect customer image data from all directions. At the same time, the system records service-related data of service personnel, such as service duration and service evaluation.

[0103] The feature extraction module is used to analyze and extract customer facial features, clothing images, facial expressions and body language features. The feature extraction module can use a variety of advanced image processing and machine learning algorithms to achieve accurate extraction of various customer features.

[0104] The big data analytics module utilizes a big data processing platform and algorithms to analyze and mine massive amounts of data, establish a rule base for association, and perform big data analysis based on customer facial features and clothing images to make preliminary predictions about customer personality and needs, thereby assisting service personnel in providing services.

[0105] The emotion analysis module is used to monitor the emotional state of customers in real time based on their facial expressions and body language characteristics, in order to assist service personnel in providing services.

[0106] The early warning analysis module is used to continuously monitor the cumulative emotional value of customers within a sliding time window, compare it with a preset emotional threshold range, and trigger different service response mechanisms based on the comparison results; intelligently adjust the maximum value of the preset emotional threshold range; and screen other service personnel when selecting other service personnel to intervene in collaborative services or take over the service work of the current customer.

[0107] The communication module uses wireless technologies such as Bluetooth and Wi-Fi to transmit information such as the preliminary prediction results of customer personality and needs from the big data analysis module, the monitoring results of customer emotional state from the emotion analysis module, and the service response instructions from the early warning analysis module to the headsets worn by service personnel in real time, so that service personnel can obtain information in a timely manner and make corresponding service strategy adjustments.

[0108] Through the above technical solution, this embodiment provides a customer demand management system based on big data analysis. The system achieves full-process management from data collection, feature extraction, demand and personality prediction, emotion monitoring to service response and personnel management through the collaborative work of various functional modules, providing strong technical support for enterprises to efficiently manage customer demand and improve service levels.

[0109] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A customer demand management method based on big data analysis, characterized in that, The method includes: S1. Collect customer image data through the acquisition module; S2. The feature extraction module analyzes and extracts the customer's facial features and clothing images. The big data analysis module uses big data analysis technology to make a preliminary prediction of the customer's personality and needs, and pushes the preliminary prediction results to the service personnel. S3. During the service process, the feature extraction module analyzes and extracts the customer's facial expressions and body language features in real time, and analyzes them in real time through the emotion analysis module to monitor the customer's emotional state in real time and assist service personnel in providing services. S4. The early warning analysis module continuously monitors the cumulative emotional value of customers within the sliding time window and compares the cumulative emotional value with the preset emotional threshold range. When the cumulative emotional value is greater than the maximum value of the preset emotional threshold range, the emotional analysis module continuously assists service personnel in providing services. When the cumulative emotional value is within the preset emotional threshold range, the emotional analysis module switches the analysis mode to manual intervention mode, and other service personnel in the store intervene to provide collaborative services. When the cumulative emotional value is less than the minimum value of the preset emotional threshold range, a service handover instruction is automatically generated to prompt other service personnel in the store to take over the service work for the current customer. In step S3, the process by which the emotion analysis module monitors the customer's emotional state in real time includes: Identify and extract customer facial expression and body language features; use formulas The emotional characteristic value R of the customer is obtained through analysis and calculation; Where N is the number of preset facial expression features. , Let M be the emotion rating value of the i-th facial expression feature, and M be the preset number of body language features. , Let j be the emotion rating value of the j-th body language feature. , This is the first weighting coefficient; A Cartesian coordinate system is established with timestamps as the X-axis and the obtained customer emotional feature values ​​R as the Y-axis. A smooth curve is then used to sequentially connect the distribution points formed by each time point and emotional feature value during the customer service process, thus forming a curve representing the customer's emotional state. ; In step S4, the process by which the early warning analysis module continuously monitors the customer's cumulative sentiment value within the sliding time window includes: The cumulative emotional value of customers within the sliding time window is obtained by analyzing and calculating using formulas (1)-(2). ; in, For sliding time windows, The curve showing how a client's negative emotional value changes over time. , This is the second weighting coefficient; Accumulated emotional value Compared with the preset emotional threshold range Perform a comparison; when At the same time, the sentiment analysis module continuously assists service personnel in providing services; when At this time, other service personnel in the store will intervene to provide assistance; when When this happens, notify other staff members in the store to take over the service for the current customer; The maximum value of the preset emotion threshold range can be intelligently adjusted in the following ways: Establish a service personnel evaluation index system to quantitatively score service personnel through three dimensions: service duration, service quality, and on-site performance. Through formula The analysis and calculation yielded the maximum value within the preset emotional threshold range. ; in, This is the base value for the preset maximum range of emotion thresholds. For adjustment coefficients, X represents the cumulative service time of the service personnel, C represents the historical service quality evaluation value of the service personnel, and C represents the on-site performance of the service personnel during the current service process. , , This is the third weighting coefficient.

2. The customer demand management method based on big data analysis according to claim 1, characterized in that, In step S2, the process by which the big data analysis module makes a preliminary prediction of customer personality based on big data analysis technology includes: Extract facial features from the customer's facial image and construct a facial feature set Q; The set of facial features Q is matched with a rule library of facial features-personality associations built based on big data analysis; If the similarity between the feature combination in Q and a certain set of facial feature combinations in the rule base reaches a preset similarity threshold, then the personality type corresponding to that set of facial feature combinations will be used as the preliminary prediction result of the customer's personality. If the similarity between the feature combination in Q and any set of facial feature combinations in the rule base does not reach the preset similarity threshold, then the personality types corresponding to the top three sets of facial feature combinations in the rule base with the highest similarity to the feature combination in Q will be used as the preliminary prediction results of the customer's personality.

3. The customer demand management method based on big data analysis according to claim 2, characterized in that, In step S2, the process by which the big data analysis module makes a preliminary prediction of customer demand based on big data analysis technology includes: Segment customer clothing images, identify clothing style, color, and material characteristics, and construct a clothing feature set W; Calculate the similarity between the clothing feature set W and each combination of clothing features in the clothing-demand association rule base established based on big data analysis; The product categories corresponding to the most similar clothing feature combinations are used as a preliminary prediction of customer demand.

4. The customer demand management method based on big data analysis according to claim 3, characterized in that, The screening process for other service personnel includes: Establish a service personnel screening index system, and quantitatively score service personnel through three dimensions: service experience, service evaluation, and service authority; Through formula The analysis and calculation yielded a comprehensive score for each service personnel. ; in, The service experience evaluation score for service personnel. The service evaluation score for service personnel The service authority evaluation value for service personnel. , , This is the fourth weighting coefficient; Service personnel are ranked from highest to lowest based on their overall scores, and those with higher overall scores are selected to provide collaborative services or take over the current customer's service work; when overall scores are the same, service personnel with higher service authority levels are given priority.

5. A customer demand management system based on big data analysis, characterized in that, The system is applied to the customer demand management method based on big data analysis as described in any one of claims 1-4, and the system includes: The data acquisition module is used to collect image data from customers and service-related data from service personnel. The feature extraction module is used to analyze and extract customer facial features, clothing images, facial expressions, and body language features; The big data analytics module is used to perform big data analysis based on customers' facial features and clothing images to make preliminary predictions about customers' personality and needs, in order to assist service personnel in providing services. The emotion analysis module is used to monitor the emotional state of customers in real time based on their facial expressions and body language characteristics, in order to assist service personnel in providing services. The early warning analysis module is used to continuously monitor the cumulative emotional value of customers within a sliding time window, compare it with a preset emotional threshold range, and trigger different service response mechanisms based on the comparison results; intelligently adjust the maximum value of the preset emotional threshold range; and screen other service personnel when selecting other service personnel to intervene in collaborative services or take over the service work of the current customer. The communication module is used to transmit in real time the preliminary prediction results of customer personality and needs from the big data analysis module, the customer emotional state monitoring results from the emotion analysis module, and the service response instructions from the early warning analysis module to the headset worn by the service personnel.

Images