Return visit tracking data management method based on big data

Abstract

The invention belongs to the technical field of analysis, application and management of return visit data, and particularly discloses and provides a return visit tracking data management method based on big data, which comprises the following steps of: layering users by adopting a clustering analysis method based on user purchase attributes, distinguishing priorities of different users in return visit investigation according to return visit application types, and determining the return visit tracking data according to the priorities; therefore, the suitability and representativeness of users with different purchase attributes in various return visit investigation scenes are ensured; according to the method, historical investigation records of users in each user layer are analyzed, the distribution effect of historical return visit modes is evaluated, and the return visit modes and return visit time of different users are subjected to differential configuration and targeted optimization according to the distribution effect, so that the response rate of the users in the return visit process is improved to the maximum extent; by evaluating the structural data stability of different user layers, the encryption storage scheme of the user layers is dynamically optimized, and the security and integrity of the user data in the storage process are ensured, so that the technical requirements of data security protection are met.

Description

Technical Field

[0001] The present invention belongs to the technical field of analysis, application and management of return visit data, and relates to a method for managing return visit tracking data based on big data. Background Art

[0002] With the rapid development and wide application of big data technology, enterprises' demand for the analysis and mining of user behavior data is increasing day by day. In the field of user relationship management, return visit tracking, as an important means to improve user satisfaction and optimize product services, the scientificity and effectiveness of its data management method directly affect the operation efficiency of enterprises and user stickiness. However, the existing technologies have obvious deficiencies in user stratification, return visit strategy optimization and data security storage, and cannot meet the refined, intelligent and security requirements of enterprises for the management of return visit tracking data. Therefore, developing a method for managing return visit tracking data based on big data has important practical significance and application value.

[0003] In the prior art, there are also some related solutions involving the analysis, application and management of return visit data. For example, the patent with the Chinese patent publication number CN116739434A discloses a service evaluation method and system based on four-dimensional customer fit management, which includes: constructing a data collection channel for evaluating customer service fit, obtaining relevant data according to each collection channel to support the calculation of fit evaluation indicators; reasonably designing a customer service return visit questionnaire in combination with the customer service business scenario; formulating a customer service return visit strategy; based on the collected customer service return visit data, unifying the acceptance message into a preset format and storing it in the customer return visit data center; constructing a comprehensive index system for evaluating customer satisfaction from four dimensions of behavior, emotion, value and strategy. It can collect and analyze customer service information from multiple dimensions, construct a customer service data center, objectively evaluate the customer service situation, and promote the continuous improvement and enhancement of enterprise service work.

[0004] Although the above solution proposes some solutions for the analysis, application and management of return visit data, there are still the following limitations: 1. The prior art fails to conduct refined stratification of user purchase attributes according to different return visit purposes of actual enterprise needs, resulting in low user response rate and insufficient resource utilization. Moreover, the analysis of users' historical return visit records is not deep enough, and it is impossible to accurately evaluate the effects of different return visit methods, making it difficult to dynamically adjust the return visit strategy, so that the return visit method lacks pertinence, and thus the return visit effect is unstable.

[0005] 2. When storing a large amount of user data, the prior art often adopts a fixed encryption storage scheme, and fails to perform dynamic optimization according to different data characteristics and stabilities, resulting in the risk of data leakage or low storage efficiency. Summary of the Invention

[0006] In view of this, to solve the problems raised in the above background art, a method for managing return visit tracking data based on big data is proposed herein.

[0007] The object of the present invention can be achieved by the following technical solutions: The present invention provides a method for managing return visit tracking data based on big data, including the following steps: In the state of setting the return visit application type, according to the historical purchase attributes, users are hierarchically prioritized through a clustering algorithm, and the return visit research behaviors of different users corresponding to each user layer are obtained, where the return visit application types include product function improvement type, customer life cycle management type, and activity promotion management type.

[0008] The response rate of the return visit research behaviors of different users corresponding to each user layer is verified to determine the distribution effect of the historical return visit methods, and the distribution effect is divided into effective and ineffective.

[0009] When the historical return visit method is effective, the response subject is marked, and the return visit plan is optimized.

[0010] When the historical return visit method is ineffective, the lost subject is marked, and the dynamic optimization setting of the return visit method is performed.

[0011] The return visit time is allocated based on the priorities of the user layers where the response subject and the lost subject are located.

[0012] The current and historical return visit data are collected, and the time series characteristics of the return visit data are evaluated through the degree of dispersion to determine the data structure stability of different user layers.

[0013] A hierarchical encryption strategy is implemented according to the data structure stability of the user layer.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Based on the user purchase attributes, the present invention uses a clustering analysis method to stratify users, and distinguishes the priorities of different users in the return visit research according to the return visit application type, so as to ensure the adaptability and representativeness of users with different purchase attributes in various return visit research scenarios.

[0015] (2) By analyzing the historical research records of users in each user layer, the present invention evaluates the distribution effect of the historical return visit methods, and accordingly makes differential configurations and targeted optimizations for the return visit methods and return visit times of different users to maximize the response rate of users during the return visit process.

[0016] (3) By evaluating the structural data stability of different user layers, the present invention dynamically optimizes the encrypted storage scheme of the user layer to ensure the security and integrity of user data during the storage process, thereby meeting the technical requirements of data security protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart of the implementation steps of the method of the present invention.

[0019] Figure 2 It is a schematic diagram of the priority stratification of users of the present invention.

[0020] Figure 3 It is a schematic flowchart of the judgment process of response and non-response of the present invention. Specific implementation manners

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0022] Please refer to Figure 1 、 2 As shown, the present invention provides a method for managing return visit tracking data based on big data, including the following steps: Tp1. In the state of setting the return visit application type, according to the historical purchase attributes, users are stratified by priority through a clustering algorithm, and the return visit research behaviors of different users in each user layer are obtained from the enterprise return visit tracking system. The return visit application types include product function improvement type, customer life cycle management type, and activity promotion management type.

[0023] The historical purchase attributes include, but are not limited to, the user's purchase records, consumption behaviors, promotion sensitivity, device usage, etc., and can be obtained by locating the user's login page through the enterprise return visit tracking system.

[0024] In a preferred implementation manner, in the state of setting the return visit application type, the step of stratifying users by priority according to the historical purchase attributes through a clustering algorithm includes: B1. Upload the return visit application type in the enterprise return visit tracking system, and match the corresponding return visit research cycle and return visit project weight distribution table of the return visit application type through the system rule library.

[0025] The system rule library is used to store the corresponding return visit research cycle, return visit project weight distribution table, research visit theme, and each question of each return visit application type.

[0026] The return visit research cycle refers to the time interval between the recording time of the recent application records in the enterprise return visit tracking system and the current return visit time.

[0027] The weight distribution table for return visit items is used to map the influence weights of different active behaviors on the return visit application types. The active behaviors are jointly composed of the user's purchase behavior, evaluation behavior, and historical participation in return visit research behavior. Among them, the participation in return visit research behavior includes the platform browsing records of the user during the research period, research inquiry topics (such as satisfaction survey topics, effect evaluation topics, improvement suggestion topics, problem follow-up topics), and the recorded voice of the user's responses during the historical research process. Each topic of the research inquiry topic corresponds to several different questions.

[0028] The product function improvement type is used to enhance the function experience and strengthen the product competitiveness; the customer life cycle management type is used to increase the customer retention rate, increase the customer's consumption amount and frequency; the activity promotion management type is used to improve the activity effect, enhance the brand influence or promote user conversion (such as converting activity participants into actual purchasing users).

[0029] The user purchase records include daily purchases, activity promotion purchases (such as the number of times of participating in promotion activities, the proportion of the amount of purchased promotional goods, etc.); the evaluation records include the evaluation text or complaint feedback text of the user on the purchased product, and the communication text between the user and the back-end customer service.

[0030] B2. Retrieve each user with active behaviors within the return visit research cycle from the enterprise return visit tracking system and mark them as each active subject. For example, when a user has a purchase behavior, an evaluation behavior, or a participation in return visit research behavior within the return visit research cycle, then mark this user as an active subject.

[0031] B3. List the characteristics of the active behaviors of each active subject and substitute them into the clustering algorithm to generate multiple clusters of active subjects with similar return visit characteristics. Take the clusters of active subjects with similar return visit characteristics as the same user layer. Each cluster of active subjects with similar return visit characteristics has similar return visit characteristics. For example, divide multiple users with the same purchase behavior or evaluation behavior into the same user layer.

[0032] In a further preferred implementation manner, the listing of the characteristics of the active behaviors of each active subject can be systematically constructed from dimensions such as the consumption behavior feature system, the text sentiment feature system, the feedback quality feature system, the promotion sensitivity feature system, and the behavior sequence feature system.

[0033] The consumption behavior feature system is used to map the return visit characteristics of the user in dimensions such as purchase frequency, consumption amount, and category preference.

[0034] The text sentiment feature system is used to map the return visit features of users in dimensions such as the positive / negative review tendency value, complaint frequency, and customer service interaction frequency.

[0035] The feedback quality feature system is used to map the return visit features of users in dimensions such as research participation, feasibility score of function improvement suggestions, and push response time difference.

[0036] The promotion sensitivity feature system is used to map the return visit features of users in dimensions such as the response rate of promotion activities, the consumption increase during the promotion period, and the promotion channel preference (such as the dwell time on the activity page).

[0037] The behavior sequence feature system is used to map the return visit features of users in dimensions such as the purchase decision chain (such as the average number of steps from browsing to purchase, the conversion rate from the promotion page to the settlement page), and the proportion of devices accessing the promotion activity (such as mobile / PC).

[0038] B4. According to the similar return visit features and the return visit item weight distribution table, judge the priority weights of each active subject cluster, and sort each active subject cluster in descending order according to the priority weights to obtain each sorted user layer.

[0039] Specifically, select from the return visit item weight distribution table the corresponding active behaviors and influence weights that match a number of similar return visit features corresponding to each active subject cluster, and accumulate the weights of all similar return visit features corresponding to each active subject cluster according to the one-to-one mapping relationship between the active behaviors and the similar return visit features to generate the priority weights of each active subject cluster.

[0040] Based on the user purchase attributes, the present invention uses a clustering analysis method to stratify users, and distinguishes the priorities of different users in the return visit research according to the return visit application type, so as to ensure the adaptability and representativeness of users with different purchase attributes in various return visit research scenarios.

[0041] Tp2. Verify the response rate of the return visit research behaviors of different users corresponding to each user layer to determine the distribution effect of the historical return visit method, and the distribution effect is divided into effective and ineffective.

[0042] In a preferred implementation manner, verifying the response rate of the return visit research behaviors of different users corresponding to each user layer to determine the distribution effect of the historical return visit method includes: obtaining from the return visit research behaviors the return visit methods and return visit research results recorded each time during the return visit research period for each active subject in each user layer, and the return visit research results are divided into having a response and having no response.

[0043] The return visit methods include email return visit, text message return visit, phone return visit, online questionnaire (communication text between the user and the background customer service), social media (the user logs in to the official websites of each platform) return visit, etc.

[0044] The response rate of each active subject in each user layer to each callback method is determined according to the callback method and callback survey results of each record. For example, the corresponding record weight of the record with a response result is set to 1, and the corresponding record weight of the record with a non-response result is set to 0; all records corresponding to different callback methods of each active subject in each user layer are clustered, and the record weights of all records are accumulated to obtain the comprehensive weight of each active subject in each user layer corresponding to each callback method; the ratio of the comprehensive weight to the preset reference weight is obtained to obtain the response rate of each active subject in each user layer to each callback method.

[0045] By comparing the response rate with the preset standard response rate, the allocation effect of each active subject in each user layer corresponding to each return visit method is determined. Specifically, if the response rate of an active subject in a certain user layer to a certain return visit method exceeds the preset standard response rate, then the allocation effect of the active subject in the user layer to the return visit method is judged to be valid; if the response rate of the active subject in the user layer to each return visit method does not exceed the preset standard response rate, then the allocation effect of the active subject in the user layer corresponding to each return visit method is judged to be invalid.

[0046] See also Figure 3 As shown, in a further preferred embodiment, the method for judging whether there is a response or not includes: setting access response rules according to the return visit method of each record, the access response rules including voice response rules and information response rules. For example, for a telephone return visit method, the access response rules are set to voice response rules; for non-telephone return visit methods such as SMS return visit and email return visit, the access response rules are set to information response rules.

[0047] The voice response rule is to track the user's reply audio. When it is recognized that the user has voice reply content after a follow-up call is made, such as it is recognized that the user has replied to one or more questions in a follow-up call, it is judged that the user's follow-up survey result in this record is responsive; otherwise, it is judged that the user's follow-up survey result in this record is unresponsive.

[0048] The information response rule is to set an access response period, such as 3 hours, 10 hours, 24 hours, etc. When it is tracked that the user has logged into the application platform to browse related products within the access response period after a certain follow-up information is released, it is judged that the user's follow-up survey result in this record is responsive; otherwise, it is judged that the user's follow-up survey result in this record is unresponsive.

[0049] Tp3. Mark the responding subject when the historical revisit method is effective, and optimize the revisit plan.

[0050] In a preferred embodiment, when the historical follow-up method is effective, mark the response subjects and optimize the follow-up plan, which includes: counting all active subjects with one or more effective follow-up methods in each user layer, marking them as the corresponding response subjects in each user layer, and selecting the item with the highest response rate among one or more effective follow-up methods as the current follow-up method for the corresponding response subjects in each user layer. For example, when there are phone follow-up and text message follow-up for a certain active subject in a certain user layer, obtain the item with the highest response rate corresponding to the phone follow-up and text message follow-up of this active subject in this user layer as the current follow-up method for this active subject in this user layer.

[0051] Match the research inquiry topics and each question corresponding to the follow-up application type through the system rule library, and identify the first-level response rate of each question belonging to the research inquiry topic corresponding to the follow-up application type among all users through natural language processing (NLP) technology. Specifically, based on the above judgment methods of having a response and not having a response, count the number of records of having a response and not having a response for each question in the corresponding follow-up research behaviors of all users. According to the formula: first-level response rate = number of records of having a response / (number of records of having a response + number of records of not having a response), obtain the first-level response rate of each question belonging to the research inquiry topic corresponding to the follow-up application type among all users.

[0052] Obtain the follow-up research behaviors of the corresponding response subjects in each user layer, screen out each question belonging to the research inquiry topic of each follow-up, and determine the second-level response rate of the corresponding response subjects in each user layer to each question belonging to the research inquiry topic corresponding to the follow-up application type. Specifically, based on the above judgment methods of having a response and not having a response, count the comprehensive research record times and the number of records of having a response of the corresponding response subjects in each user layer to each question belonging to the research inquiry topic corresponding to the follow-up application type. According to the formula: second-level response rate = number of records of having a response / comprehensive research record times, obtain the second-level response rate of the corresponding response subjects in each user layer to each question belonging to the research inquiry topic corresponding to the follow-up application type.

[0053] Accumulate the first-level response rate and the second-level response rate to generate the comprehensive response rate of the corresponding response subjects in each user layer to each question belonging to the research inquiry topic corresponding to the follow-up application type, and replace the questions with a comprehensive response rate lower than the preset standard response rate.

[0054] Tp4. When the historical follow-up method is ineffective, mark the lost subjects and perform dynamic optimization settings for the follow-up method.

[0055] In a preferred embodiment, when the historical follow-up method is ineffective, mark the lost subjects and perform dynamic optimization settings for the follow-up method, which includes: marking all active subjects other than the corresponding response subjects in each user layer as the corresponding lost subjects in each user layer, and integrating multiple follow-up methods recorded by the corresponding lost subjects in each user layer.

[0056] For each lost entity in each user layer where the follow-up method is not recorded, select any one of the unrecorded follow-up methods as the current follow-up method. For example, for a lost entity in a certain user layer that has not recorded the email follow-up method and the social media follow-up method, randomly select one from the email follow-up method and the social media follow-up method as the current follow-up method for this lost entity in this user layer.

[0057] For each lost entity in each user layer where all follow-up methods are recorded, formulate preferential policies according to the priority weights to promote user follow-up response. For example, for the lost entities in the first user layer, set multiple follow-up methods in coordination, and when setting the corresponding access response rules for these follow-up methods, distribute a certain amount of coupons.

[0058] Tp5. Allocate follow-up time based on the priorities of the user layers where the response entities and the lost entities are located.

[0059] In a preferred implementation manner, the allocation of follow-up time based on the priorities of the user layers where the response entities and the lost entities are located includes: performing time alignment processing on the historical purchase attributes of the corresponding active entities in each user layer in a unified time period manner to determine the response probabilities of the corresponding active entities in each user layer in each time period. Specifically, each working day is divided into different time stages, such as 2:00 - 5:00 and 5:00 - 8:00, etc. By identifying the comprehensive record times of each corresponding active entity in each user layer logging in to the platform in different time stages through the historical purchase attributes, obtaining the ratio of the comprehensive record times to the preset reference comprehensive record times, and obtaining the response probabilities of the corresponding active entities in each user layer in each time period.

[0060] Select the time period with the highest response probability as the follow-up information release time period for the corresponding response entities and lost entities in each user layer.

[0061] Count the current follow-up methods of the corresponding response entities and lost entities in each user layer, and screen out the entities whose current follow-up method is telephone follow-up among the response entities and the lost entities, and mark them as the corresponding inquiry entities in each user layer.

[0062] Count different entities in the same follow-up information release time period, perform a priority ranking for telephone follow-up on different entities in the same follow-up information release time period according to the user layer priorities, and then conduct telephone follow-up on different entities in different follow-up information release time periods in sequence according to the order of this telephone follow-up priority ranking.

[0063] The present invention analyzes the historical research records of users in each user layer, evaluates the distribution effect of the historical return visit method, and accordingly makes differential configuration and targeted optimization of the return visit methods and return visit times for different users to maximize the response rate of users during the return visit process.

[0064] Tp6. Collect current and historical return visit data, and evaluate the time series characteristics of the return visit data through the degree of dispersion to determine the data structure stability of different user layers.

[0065] The return visit data refers to the data of corresponding user changes and historical purchase attribute changes in each user layer during different return visit research behaviors.

[0066] In a preferred embodiment, the collecting of current and historical return visit data, evaluating the time series characteristics of the return visit data through the degree of dispersion to determine the data structure stability of different user layers includes: using statistical analysis tools (such as the pandas, NumPy libraries in Python and the R language environment, etc.) to organize the return visit data of each user layer into a time series data set according to the user identification and time dimension, and conducting a degree of dispersion evaluation for the time series data set of each user layer. The time series data set consists of multiple features in the active behaviors recorded by users at different times. Specifically, compare the ratio of the corresponding comprehensive change amplitude of multiple features recorded by different user layers at different times to the preset reference amplitude. The greater the change amplitude, the greater the degree of dispersion.

[0067] The change amplitude includes but is not limited to: for the same user, the corresponding change amplitude of the cumulative comprehensive purchase amount of different products; for different users, the cumulative comprehensive purchase amount of these two or more users, and compare to generate the comprehensive product change amplitude between different users.

[0068] The user identification is such as the identity name number, etc.

[0069] Conduct stability level division according to different degree of dispersion intervals. For example, the degree of dispersion within the first degree of dispersion interval is classified as high stability, the degree of dispersion within the second degree of dispersion interval is classified as medium stability, and the degree of dispersion within the third degree of dispersion interval is classified as low stability. Match the stability of the corresponding degree of dispersion of the time series data set of each user layer. The stability is divided into high stability, medium stability, and low stability. The stability level is used to map the data structure relationship of the user layer.

[0070] Tp7. Implement a hierarchical encryption strategy according to the data structure stability of the user layer.

[0071] In a preferred embodiment, the hierarchical encryption policy is implemented according to the data structure stability of the user layer, specifically: lightweight encryption is used for the user layer with high stability, and high-strength encryption protection is implemented for the user layer with low stability.

[0072] For example: For the user layer with high stability, the AES-128-bit encryption algorithm is used for encrypted storage. The AES-128-bit encryption algorithm has high encryption efficiency and can quickly complete encryption and decryption operations while ensuring data security.

[0073] For the user layer with medium stability, the AES-256-bit encryption algorithm with higher security is used. Although the encryption and decryption speed of this algorithm is slightly slower than that of the AES-256-bit algorithm, it can provide stronger encryption strength to cope with the possible security risks in the medium-stable state of the data structure.

[0074] For the user layer with low stability, in addition to using the AES-256-bit encryption algorithm, an additional layer of data obfuscation processing is added. Data obfuscation increases the complexity of the data further before encryption by performing specific transformations and recombinations on the original data to improve data security.

[0075] The present invention dynamically optimizes the encrypted storage scheme of the user layer by evaluating the structural data stability of different user layers, ensuring the security and integrity of user data during storage, thereby meeting the technical requirements of data security protection.

[0076] The above content is only an example and explanation of the concept of the present invention. Those skilled in the art of the present technology can make various modifications or supplements to the described specific embodiments or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A return visit tracking data management method based on big data, characterized in that: The steps include: When the return visit application type is set, users are prioritized and stratified according to their historical purchase attributes through a clustering algorithm to obtain return visit research behaviors of different users in each user layer. The return visit application types include product function improvement type, customer life cycle management type, and activity promotion management type. Verify the response rate of the return visit survey behavior of different users in each user layer to determine the distribution effect of the historical return visit method, and the distribution effect is divided into effective and ineffective; Mark the responding subject when the historical revisit method is effective, and optimize the revisit plan; If the historical revisit method is invalid, mark the lost subjects and perform dynamic optimization settings for the revisit method; Allocate return visit time based on the priority of the user layer where the responding subject and the churned subject are located; Collect current and historical return visit data, and evaluate the time series characteristics of the return visit data by the degree of dispersion to determine the stability of the data structure for different user layers; Implement a hierarchical encryption strategy based on the data structure stability of the user layer.

2. A return visit tracking data management method based on big data according to claim 1, characterized in that: In the state of setting the return application type, the user is prioritized according to the historical purchase attributes through the clustering algorithm, including: uploading the return application type in the enterprise return tracking system, matching the return survey cycle and the return project weight distribution table corresponding to the return application type through the system rule library; The return visit item weight allocation table is used to map the influence weights of different active behaviors on the return visit application types, wherein the active behaviors are composed of the user's purchase behavior, evaluation behavior, and historical return visit research behavior, wherein the return visit research behavior includes the user's platform reading records during the research period, the research inquiry topic, and the user's reply voice record during the historical research process; Retrieve users who have active behaviors during the return visit and survey period from the enterprise return visit tracking system and mark them as active subjects; The active behaviors of each active subject are characterized and substituted into the clustering algorithm to generate multiple active subject clusters with similar return visit characteristics. The active subject clusters with similar return visit characteristics are regarded as the same user layer, and each active subject cluster with similar return visit characteristics has similar return visit characteristics. According to the similar return visit characteristics and the return visit item weight distribution table, the priority weight of each active subject cluster is judged, and the active subject clusters are sorted in descending order according to the priority weight to obtain the sorted user layers.

3. A return visit tracking data management method based on big data according to claim 2, characterized in that: The characteristic listing of active behaviors of each active subject can be systematically constructed from the dimensions of consumption behavior characteristic system, text emotion characteristic system, feedback quality characteristic system, promotion sensitivity characteristic system, and behavior sequence characteristic system.

4. The method for managing return visit tracking data based on big data according to claim 2, characterized in that: The response rate verification of the return visit survey behavior of different users in each user layer is performed to determine the historical return visit method distribution effect, including: obtaining the return visit method and return visit survey results of each active subject in each user layer during the return visit survey cycle from the return visit survey behavior, and the return visit survey results are divided into response and non-response; Determine the response rate of each active subject in each user layer to each return visit method according to the return visit method and return visit survey results recorded each time; By comparing the response rate with the preset standard response rate, the allocation effect of each return visit method for each active subject in each user layer is determined.

5. A return visit tracking data management method based on big data according to claim 4, characterized in that: The judgment method of response and non-response includes: setting access response rules according to the return visit method of each record, and the access response rules include voice response rules and information response rules; The voice response rule is to track the user's reply audio. When it is recognized that the user has a voice reply content after a certain callback call is made, it is determined that the user's callback survey result recorded at that time is responsive; otherwise, it is determined that the user's callback survey result recorded at that time is unresponsive; The information response rule is to set an access response period. When it is tracked that the user has logged into the application platform to browse related products within the access response period after a certain follow-up information is released, it is judged that the user's follow-up survey result in this record is responsive; otherwise, it is judged that the user's follow-up survey result in this record is unresponsive.

6. A return visit tracking data management method based on big data according to claim 4, characterized in that: The marking of the responding subject when the historical revisiting method is valid and the optimization of the revisiting scheme include: counting the active subjects in each user layer with one or more valid revisiting methods, marking them as the responding subjects corresponding to each user layer, and selecting the one with the highest response rate among the one or more valid revisiting methods as the current revisiting method of each responding subject corresponding to each user layer; Match the survey inquiry topics and questions corresponding to the application type for the follow-up visit through the system rule library, and identify the first-level response rate of all users for each question belonging to the survey inquiry topic corresponding to the application type for the follow-up visit through natural language processing technology; Obtain the return visit survey behavior of each responding subject corresponding to each user layer, screen out the questions belonging to the survey inquiry theme of each return visit, and determine the secondary response rate of each responding subject corresponding to each user layer to the questions belonging to the survey inquiry theme corresponding to the return visit application type; The first-level response rate and the second-level response rate are accumulated to generate the comprehensive response rate of each responding subject in each user layer to each question belonging to the corresponding survey inquiry topic of the follow-up application type, and the questions with a comprehensive response rate lower than the preset standard response rate are replaced.

7. A return visit tracking data management method based on big data according to claim 6, characterized in that: The marking of lost subjects when the historical revisiting method is invalid and the dynamic optimization setting of the revisiting method include: marking each active subject except each responding subject in each user layer as each lost subject corresponding to each user layer, integrating multiple revisiting methods recorded by each lost subject corresponding to each user layer; For each churn subject in each user layer whose return visit method is not recorded, select any one of the one or more return visit methods that have not been recorded as the current return visit method; For each churn subject in each user layer where all return visit methods are recorded, preferential policies are formulated according to priority weights to promote user return visit responses.

8. The method for managing return visit tracking data based on big data according to claim 7, characterized in that: The revisit time allocation based on the priority of the user layer where the responding subject and the churned subject are located includes: using a unified time period to perform time alignment processing on the historical purchase attributes of each active subject corresponding to each user layer to determine the response probability of each active subject corresponding to each user layer in each time period; The time period with the highest response probability is selected as the time period for publishing the return visit information of each responding subject and each churn subject corresponding to each user layer; The current return visit methods of each responding subject and each churn subject corresponding to each user layer are counted, and the subjects whose current return visit method is telephone return visit among each responding subject and each churn subject are screened out, and marked as the corresponding inquiry subject of each user layer; Count the different subjects in the same follow-up information release period, prioritize the different subjects in the same follow-up information release period for telephone follow-up according to the user layer priority, and then conduct telephone follow-up on the different subjects in different follow-up information release periods in turn according to the telephone follow-up priority order.

9. The method for managing return visit tracking data based on big data according to claim 3, characterized in that: The collecting of current and historical return visit data, and evaluating the time series characteristics of the return visit data by the degree of dispersion, so as to determine the stability of the data structure of different user layers, includes: using statistical analysis tools to organize the return visit data of each user layer into a time series data set according to user identification and time dimension, and conducting a degree of dispersion evaluation on the time series data set of each user layer, wherein the time series data set is composed of multiple features of the active behaviors of users recorded at different times; The stability level is divided according to different discrete degree intervals to match the stability of the corresponding discrete degree of the time series data set of each user layer. The stability is divided into high stability, medium stability, and low stability. The stability level is used to map the data structure relationship of the user layer.

10. A return visit tracking data management method based on big data according to claim 9, characterized in that: The hierarchical encryption strategy is implemented according to the stability of the user layer, specifically: lightweight encryption is used for the user layer with high stability, and high-intensity encryption protection is implemented for the user layer with low stability.

Images