Market management method and system based on RSC

Abstract

The invention relates to the technical field of market management, in particular to an RSC-based market management method and system. The method comprises the following steps: acquiring multi-source market data and user interaction data by using RSC; analyzing user interaction data, and determining a function dependency degree; determining a user viscosity change according to the function dependency degree; analyzing the multi-source market data according to the user viscosity change, and predicting a user loss risk level and a user loss direction; and generating a market analysis report according to the user loss risk level and the user loss direction. The dynamic evaluation of the viscosity change of the user is realized, so that the evaluation is not limited to static snapshot any more, and the deviation caused by ignoring the dynamic restriction of an external market is avoided. Multi-dimensional prediction of the user loss risk is realized, the defect of dependence on a single data source is avoided, and incomplete prediction caused by functional defects of unprocessed products is avoided at the same time. The timeliness and accuracy of decision making are improved, and invalid reports caused by prediction deviation are avoided.

Description

Technical Field

[0001] The present application relates to the technical field of market management, and in particular to a market management method and system based on RSC. Background Art

[0002] In the current market management landscape, with the prevalence of data-driven decision-making, analytical methods based on data platforms have become a core means of improving user engagement and predicting user churn. As an advanced data processing system, the RSC platform has demonstrated significant advantages in data fusion, intelligent analysis, and trend insights.

[0003] However, in the process of quantitative evaluation of user stickiness, the existing RSC-based market management method lacks a systematic quantification mechanism, which makes the evaluation of user stickiness changes too static, resulting in inaccurate analysis of user stickiness changes and hindering the formulation of precise user churn intervention strategies. Summary of the Invention

[0004] This application provides a market management method and system based on RSC to solve the above problems.

[0005] In a first aspect, the present application provides a market management method based on RSC, the method comprising: Use RSC to obtain multi-source market data and user interaction data; Analyzing the user interaction data to determine functional dependencies; Determining a change in user stickiness based on the functional dependency; Analyze the multi-source market data based on the changes in user stickiness to predict the user churn risk level and user churn direction; A market analysis report is generated based on the user churn risk level and the user churn direction.

[0006] This solution utilizes RSC to obtain multi-source market data and user interaction data, avoiding evaluation biases caused by data delays. Analyze user interaction data to determine functional dependencies, ensure the accuracy of user stickiness change assessments, and avoid chain reactions caused by dependency errors. Based on functional dependencies, determine changes in user stickiness, and implement dynamic evaluations of user stickiness changes, so that evaluations are no longer limited to static snapshots and avoid biases caused by ignoring dynamic external market constraints. Based on user stickiness changes, analyze multi-source market data to predict the level of user churn risk and the direction of user churn, achieving multi-dimensional predictions of user churn risk, avoiding the shortcomings of relying on a single data source, and avoiding incomplete predictions caused by unaddressed product functional defects. Generate market analysis reports based on the level of user churn risk and the direction of user churn, improving the timeliness and accuracy of decision-making and avoiding invalid reports caused by prediction bias.

[0007] Optionally, analyzing the user interaction data to determine the functional dependency includes: Parsing the user interaction data to determine user operation data and response interaction data; Analyze the user operation data to determine the login frequency and function usage depth; Determining the intensity of function usage based on the login frequency and the depth of function usage; Analyze the response interaction data to determine the API call volume and transaction processing time; Determine the function response efficiency based on the API call volume and the transaction processing time; The function dependency is determined according to the function usage intensity and the function response efficiency.

[0008] Through this solution, user interaction data is parsed, user operation data and response interaction data are determined, and a clear division of data sources is achieved to avoid quantitative deviations caused by data confusion. User operation data is analyzed to determine the login frequency and function usage depth, effectively supporting the quantitative evaluation of user active usage behavior. Based on the login frequency and function usage depth, the function usage intensity is determined to characterize the user's active dependence on the function and eliminate the deviation of a single indicator. Response interaction data is analyzed to determine the API call volume and transaction processing time, and the original interaction log is converted into a computable numerical indicator to provide input for the calculation of function response efficiency. Based on the API call volume and transaction processing time, the function response efficiency is determined, the efficiency evaluation is simplified, and a side input is provided for the calculation of function dependency. Based on the function usage intensity and function response efficiency, the function dependency is determined to quantify the user's dependence on the platform function.

[0009] Optionally, determining the function dependency according to the function usage intensity and the function response efficiency includes: Analyze the multi-source market data to determine competing product data; Analyze the competitor data to determine the functional coverage of the competitor; Determine the impact of functional dependencies based on the functional coverage of the competitor product; The function dependency is determined according to the function dependency impact, the function usage intensity and the function response efficiency.

[0010] This solution parses multi-source market data, identifies competitor data, and ensures that external market factors are effectively isolated and prepared, avoiding the inability to fully analyze the interactions between data sources. Competitive data is analyzed to determine the coverage of competitor functions, ensuring that external competitive factors are included in the assessment and eliminating the problem of ignoring the dynamic constraints of competitors. Based on the coverage of competitor functions, the impact of functional dependencies is determined, ensuring that multi-source data is integrated to improve the accuracy of functional dependency quantification. Based on the impact of functional dependencies, functional usage intensity, and functional response efficiency, the degree of functional dependency is determined, eliminating the problem of being unable to fully analyze the interactions between data sources and providing reliable input for user stickiness assessment.

[0011] Optionally, determining a change in user stickiness based on the functional dependency includes: Analyzing a changing trend of the functional dependency over time to determine a functional dependency decay rate; Analyzing the user interaction data to determine user activity period data; Determining user behavior stability based on the user activity cycle data; The user stickiness change is determined based on the function dependency decay rate and the user behavior stability.

[0012] This solution analyzes the changing trends of functional dependencies over time, determines the functional dependency decay rate, ensures the evolution of functional dependencies is quantified, and supports dynamic processing in the RSC platform's streaming computing engine, thereby enhancing the accuracy of user stickiness assessments. It also analyzes user interaction data to determine user activity cycle data, improving the basis for assessing user loyalty fluctuations and avoiding static biases. Based on user activity cycle data, it determines user behavior stability, ensures behavioral stability is quantified, and supports dynamic adjustments to user stickiness changes. Based on the functional dependency decay rate and user behavior stability, it determines user stickiness changes, ensuring that user stickiness change assessments integrate multi-source data in real time, thereby enhancing the timeliness of market decisions.

[0013] Optionally, determining a change in user stickiness based on the function dependency decay rate and the user behavior stability includes: Analyze the multi-source market data to determine industry trend data; Analyze the industry trend data to determine the market volatility index; The change in user stickiness is determined based on the market volatility index, the function dependency decay rate, and the user behavior stability.

[0014] This solution analyzes multi-source market data to identify industry trend data, avoiding the inability to fully analyze the interplay between data sources and ensuring that user stickiness change assessments are based on integrated multi-source data. By analyzing industry trend data and determining the market volatility index, we improve the accuracy and timeliness of user stickiness change predictions and support market decision-making. Based on the market volatility index, feature dependency decay rate, and user behavior stability, we determine user stickiness changes, improve the effectiveness of market analysis reports, and provide a basis for data-driven decision-making.

[0015] Optionally, analyzing the multi-source market data based on the change in user stickiness to predict the user churn risk level and user churn direction includes: Comparing the user stickiness change with a preset stickiness threshold, and determining a low stickiness level where the user stickiness is lower than the preset stickiness threshold; Analyze multi-source market data corresponding to the aforementioned low stickiness to determine market positioning; Determine the direction of user churn based on the market positioning; Analyze the change difference between the low stickiness level and a preset stickiness threshold; and determine the user churn risk level based on the change difference.

[0016] This solution compares changes in user stickiness with preset stickiness thresholds, identifying low stickiness levels where user stickiness falls below the threshold. This quantifies whether user loyalty has reached a critical low point, ensuring that only cases with insufficient stickiness are further addressed, avoiding wasting resources on users with normal stickiness. Multi-source market data corresponding to low stickiness levels is analyzed to determine market positioning and reveal the external drivers of low stickiness, providing context for identifying the market root causes of declining user loyalty. Based on market positioning, the direction of user churn is determined, providing clear directional information on user churn, effectively focusing on targeted interventions, and thus improving the practicality of predictions. The difference in change between the low stickiness level and the preset stickiness threshold is analyzed to objectively measure the depth of insufficient stickiness and avoid subjective bias. Based on this difference in change, the user churn risk level is determined and the probability of user churn is quantified.

[0017] Optionally, determining the direction of user churn based on the market positioning includes: Analyze the functional coverage of the competing products and determine the functional coverage between the market positioning and the competing products; parsing the user interaction data to determine transaction processing exception data; Analyze the transaction processing exception data to determine product functional defects; Determine the direction of user loss based on the functional coverage and product functional defects.

[0018] This solution analyzes the functional coverage of competing products, determines the functional coverage of the platform compared to competitors in terms of market positioning, and identifies the platform's relative weaknesses in functional coverage. It also analyzes user interaction data to identify transaction anomalies, improving the real-time and accuracy of data processing. Transaction anomaly data analysis identifies product functional defects, providing root cause information for determining user churn. Based on functional coverage and product functional defects, user churn trends are determined, supporting churn risk prediction decisions.

[0019] Optionally, determining the user churn risk level according to the change difference includes: Obtaining historical user churn data, analyzing the historical user churn data, and determining a churn risk baseline value; Analyze the multi-source market data to identify changes in industry policies; Determine the policy impact coefficient based on the industry policy changes; The user churn risk level is determined based on a weighted calculation of the change difference, the churn risk baseline value, and the policy impact coefficient.

[0020] This solution captures and analyzes historical user churn data to determine a churn risk baseline. This baseline serves as a quantitative benchmark for user churn risk prediction and is used as a benchmark input in weighted calculations. Multi-source market data is analyzed to identify industry policy changes, ensuring that external market factors are incorporated into risk analysis. Based on industry policy changes, a policy impact coefficient is determined to represent the positive or negative impact of the policy change on risk, which is used to adjust weights in weighted calculations. A weighted calculation of the change difference, the churn risk baseline, and the policy impact coefficient determines the user churn risk level, providing a quantitative indication of the probability of user churn and serving as a clear risk basis for market decision-making.

[0021] Optionally, analyzing the transaction processing exception data to determine product functional defects includes: Analyzing the transaction processing exception data to determine the abnormal temporal and spatial distribution; Determining abnormal aggregation characteristics based on the abnormal spatiotemporal distribution; Acquire system architecture topology data, and locate abnormal related modules based on the system architecture topology data; Matching the abnormal aggregation feature with the abnormal association module to obtain a matching relationship; determining a defect propagation path according to the matching relationship; Determine product functional defects based on the defect propagation path.

[0022] This solution analyzes transaction processing anomaly data, determines the spatiotemporal distribution of anomalies, and effectively reveals the spatiotemporal hotspots where anomalies occur, avoiding one-sided analysis based on a single dimension. Based on the spatiotemporal distribution of anomalies, determine anomaly aggregation characteristics, and quantify the concentration and intensity characteristics of anomalies. Obtain system architecture topology data, locate anomaly-related modules based on the system architecture topology data, narrow the scope of the problem, and ensure that the focus is on the module where the anomaly actually occurs. Match the anomaly aggregation characteristics with the anomaly-related modules to obtain a matching relationship, and confirm evidence of anomalies in different modules under spatiotemporal conditions. Based on the matching relationship, determine the defect propagation path, reveal the dynamic propagation mechanism of the defect, and provide visual path data for identifying the source and diffusion point of the defect. Based on the defect propagation path, determine product functional defects, clarify the nature and impact of the defects, and improve the accuracy of user stickiness prediction.

[0023] In a second aspect, the present application provides a market management system based on RSC, the system comprising: Data acquisition module, used to obtain multi-source market data and user interaction data using RSC; An interaction data analysis module, configured to analyze the user interaction data and determine functional dependencies; A stickiness analysis module, configured to determine changes in user stickiness based on the functional dependency; A churn analysis module, configured to analyze the multi-source market data based on the changes in user stickiness, and predict the user churn risk level and user churn direction; The report generation module is used to generate a market analysis report based on the user churn risk level and the user churn direction.

[0024] Optionally, when analyzing the user interaction data to determine the functional dependency, the interaction data analysis module is used to: Parsing the user interaction data to determine user operation data and response interaction data; Analyze the user operation data to determine the login frequency and function usage depth; Determining the intensity of function usage based on the login frequency and the depth of function usage; Analyze the response interaction data to determine the API call volume and transaction processing time; Determine the function response efficiency based on the API call volume and the transaction processing time; The function dependency is determined according to the function usage intensity and the function response efficiency.

[0025] Optionally, when determining the function dependency based on the function usage intensity and the function response efficiency, the interaction data analysis module is configured to: Analyze the multi-source market data to determine competing product data; Analyze the competitor data to determine the functional coverage of the competitor; Determine the impact of functional dependencies based on the functional coverage of the competitor product; The function dependency is determined according to the function dependency impact, the function usage intensity and the function response efficiency.

[0026] Optionally, when the stickiness analysis module determines a change in user stickiness based on the functional dependency, it is configured to: Analyzing a changing trend of the functional dependency over time to determine a functional dependency decay rate; Analyzing the user interaction data to determine user activity period data; Determining user behavior stability based on the user activity cycle data; The user stickiness change is determined based on the function dependency decay rate and the user behavior stability.

[0027] Optionally, when determining a change in user stickiness based on the function dependency decay rate and the user behavior stability, the stickiness analysis module is configured to: Analyze the multi-source market data to determine industry trend data; Analyze the industry trend data to determine the market volatility index; The change in user stickiness is determined based on the market volatility index, the function dependency decay rate, and the user behavior stability.

[0028] Optionally, when the churn analysis module analyzes the multi-source market data based on the change in user stickiness and predicts the user churn risk level and user churn direction, it is configured to: Comparing the user stickiness change with a preset stickiness threshold, and determining a low stickiness level where the user stickiness is lower than the preset stickiness threshold; Analyze multi-source market data corresponding to the aforementioned low stickiness to determine market positioning; Determine the direction of user churn based on the market positioning; Analyze the change difference between the low stickiness level and a preset stickiness threshold; and determine the user churn risk level based on the change difference.

[0029] Optionally, when the churn analysis module determines the direction of user churn based on the market positioning, it is configured to: Analyze the functional coverage of the competing products and determine the functional coverage between the market positioning and the competing products; parsing the user interaction data to determine transaction processing exception data; Analyze the transaction processing exception data to determine product functional defects; Determine the direction of user loss based on the functional coverage and product functional defects.

[0030] Optionally, when the churn analysis module determines the user churn risk level based on the change difference, it is configured to: Obtaining historical user churn data, analyzing the historical user churn data, and determining a churn risk baseline value; Analyze the multi-source market data to identify changes in industry policies; Determine the policy impact coefficient based on the industry policy changes; The user churn risk level is determined based on a weighted calculation of the change difference, the churn risk baseline value, and the policy impact coefficient.

[0031] Optionally, when the churn analysis module analyzes the transaction processing exception data and determines a product function defect, it is used to: Analyzing the transaction processing exception data to determine the abnormal temporal and spatial distribution; Determining abnormal aggregation characteristics based on the abnormal spatiotemporal distribution; Acquire system architecture topology data, and locate abnormal related modules based on the system architecture topology data; Matching the abnormal aggregation feature with the abnormal association module to obtain a matching relationship; determining a defect propagation path according to the matching relationship; Determine product functional defects based on the defect propagation path. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0033] Figure 1 A schematic diagram of an application scenario provided in one embodiment of the present application; Figure 2 A flowchart of a market management method based on RSC provided in one embodiment of the present application; Figure 3 A schematic diagram of the structure of an RSC-based market management system provided in one embodiment of the present application. DETAILED DESCRIPTION

[0034] To make the purpose, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0035] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates an "or" relationship between the related objects.

[0036] The embodiments of the present application are described in further detail below with reference to the accompanying drawings.

[0037] In the process of quantitative evaluation of user stickiness, the existing RSC-based market management method lacks a systematic quantification mechanism, which makes the evaluation of user stickiness changes too static, resulting in inaccurate analysis of user stickiness changes and hindering the formulation of precise user churn intervention strategies.

[0038] Based on this, the present application provides a market management method and system based on RSC, which uses RSC to obtain multi-source market data and user interaction data to avoid evaluation deviations caused by data delays. Analyze user interaction data, determine functional dependencies, ensure the accuracy of user stickiness change evaluations, and avoid chain problems caused by dependency errors. According to functional dependencies, determine user stickiness changes, and implement dynamic evaluation of user stickiness changes, so that the evaluation is no longer limited to static snapshots, avoiding deviations caused by ignoring external market dynamic constraints. According to user stickiness changes, analyze multi-source market data, predict user churn risk levels and user churn directions, and implement multi-dimensional predictions of user churn risks, avoiding the shortcomings of relying on a single data source, and avoiding incomplete predictions caused by unaddressed product functional defects. Generate market analysis reports based on user churn risk levels and user churn directions to improve the timeliness and accuracy of decision-making and avoid invalid reports caused by prediction deviations.

[0039] Figure 1 This is a schematic diagram of an application scenario provided by this application. When conducting market management, the method provided by this application is applied.

[0040] Specifically, the method provided in this application is applied to any server, and the server interacts with user devices and online media platforms. Using RSC, user interaction data is obtained through user devices, and multi-source market data is obtained through online media platforms. Analyze user interaction data to determine functional dependencies. Based on functional dependencies, determine changes in user stickiness. Based on changes in user stickiness, analyze multi-source market data, predict the user churn risk level and user churn direction, and achieve multi-dimensional prediction of user churn risk, avoiding the shortcomings of relying on a single data source, and avoiding incomplete predictions caused by unaddressed product functional defects. Generate a market analysis report based on the user churn risk level and user churn direction to improve the timeliness and accuracy of decision-making and avoid invalid reports caused by prediction bias.

[0041] For specific implementation methods, please refer to the following embodiments.

[0042] Figure 2 This is a flow chart of a market management method based on RSC provided in one embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201. Using RSC, obtain multi-source market data and user interaction data; RSC can be a platform for dynamically processing heterogeneous data streams based on demand, market, and competition-based streaming computing engines.

[0043] Multi-source market data can be external market data streams such as industry trend reports, competitor dynamics, industry white papers, and competitor API documents.

[0044] User interaction data can be user operation logs and business system interaction data.

[0045] Specifically, the server is connected to the online media platform through the API interface. At the same time, the server has a built-in log collection module and a monitoring module, which acquire and integrate data in real time through the RSC streaming computing engine: first, the external market data obtained from the online media platform based on the RSC real-time access API interface is used to generate multi-source market data; secondly, the log collection module records the user's operation events on the business platform to generate user operation logs; at the same time, the monitoring module captures business device interaction events to generate business system interaction data; finally, the user operation log and business system interaction data constitute user interaction data.

[0046] S202, analyzing user interaction data to determine functional dependencies; Functional dependency can be the degree to which users rely on platform functions.

[0047] Specifically, the login frequency and function usage depth are extracted from the user operation log; at the same time, the API call volume and transaction processing time are extracted from the business system interaction data; then, based on each user, the login frequency, function usage depth, API call volume and transaction processing time are comprehensively considered, and the function dependency is determined through a weighted average algorithm.

[0048] S203: Determine user stickiness changes based on functional dependencies; Changes in user stickiness can be fluctuations in user loyalty to the platform.

[0049] Specifically, user behavior stability data is constructed based on the variance analysis of the behavioral pattern fluctuations of user interaction data within a preset time window; then, a time series analysis of functional dependencies is performed; and further, combined with the user behavior stability data, the changes in user stickiness are determined through the rule engine.

[0050] S204. Analyze multi-source market data based on changes in user stickiness to predict the user churn risk level and user churn direction; The user churn risk level may reflect the probability of user churn.

[0051] The direction of user churn can be the path where users turn to competing products.

[0052] Specifically, we analyze multi-source market data to extract the market volatility index and functional coverage of competitor dynamics; then, combined with changes in user stickiness, we group users through decision tree rules to predict the level of user churn risk; subsequently, we analyze functional coverage to determine the possibility of users switching to competitors; at the same time, we identify product functional defects and determine the direction of user churn.

[0053] S205: Generate a market analysis report based on the user churn risk level and user churn direction.

[0054] The market analysis report can be a document containing prediction results such as the user churn risk level and the user churn direction.

[0055] Specifically, the user churn risk level and user churn direction are integrated; then, text and charts are automatically compiled to generate a market analysis report.

[0056] This solution utilizes RSC to obtain multi-source market data and user interaction data, avoiding evaluation biases caused by data delays. Analyze user interaction data to determine functional dependencies, ensure the accuracy of user stickiness change assessments, and avoid chain reactions caused by dependency errors. Based on functional dependencies, determine changes in user stickiness, and implement dynamic evaluations of user stickiness changes, so that evaluations are no longer limited to static snapshots and avoid biases caused by ignoring dynamic external market constraints. Based on user stickiness changes, analyze multi-source market data to predict the level of user churn risk and the direction of user churn, achieving multi-dimensional predictions of user churn risk, avoiding the shortcomings of relying on a single data source, and avoiding incomplete predictions caused by unaddressed product functional defects. Generate market analysis reports based on the level of user churn risk and the direction of user churn, improving the timeliness and accuracy of decision-making and avoiding invalid reports caused by prediction bias.

[0057] In some embodiments, user interaction data is parsed to determine user operation data and response interaction data; user operation data is analyzed to determine login frequency and function usage depth; function usage intensity is determined based on login frequency and function usage depth; response interaction data is analyzed to determine API call volume and transaction processing time; function response efficiency is determined based on API call volume and transaction processing time; function dependency is determined based on function usage intensity and function response efficiency.

[0058] The user operation data may be data extracted from a user operation log.

[0059] The response interaction data may be data extracted from the business system interaction data.

[0060] Login frequency can be the number of login events that occur when a user accesses the platform.

[0061] Function usage depth can be the degree to which users use platform functions in depth.

[0062] Function usage intensity can be a comprehensive indicator that represents the degree of active dependence of users on functions.

[0063] The API call volume can be the number of API call events.

[0064] The transaction processing duration may be the length of time it takes to process a transaction.

[0065] Functional response efficiency can be a measure of the response performance to user requests.

[0066] Specifically, user interaction data is parsed, and user operation logs are extracted as user operation data through data type matching. Simultaneously, business system interaction data is extracted as response interaction data. Then, based on the user operation data, an event counter is used to count the number of login events in the user operation data to determine login frequency. Simultaneously, an indicator evaluator is used to analyze the function usage records in the user operation data to calculate function usage depth. Login frequency and function usage depth are then input into a comprehensive calculation model, where function usage intensity is generated through normalization and weighted averaging. Then, based on the response interaction data, an event counter is used to count the number of API call events in the response interaction data to determine API call volume. Simultaneously, a time meter is used to analyze the transaction records in the response interaction data to calculate transaction duration. Finally, API call volume and transaction duration are input into a comprehensive calculation model, where efficiency conversion and comprehensive scoring are performed to generate function response efficiency. Finally, function usage intensity and function response efficiency are integrated and fused through comprehensive weighting and normalization to determine function dependency.

[0067] Through this solution, user interaction data is parsed, user operation data and response interaction data are determined, and a clear division of data sources is achieved to avoid quantitative deviations caused by data confusion. User operation data is analyzed to determine the login frequency and function usage depth, effectively supporting the quantitative evaluation of user active usage behavior. Based on the login frequency and function usage depth, the function usage intensity is determined to characterize the user's active dependence on the function and eliminate the deviation of a single indicator. Response interaction data is analyzed to determine the API call volume and transaction processing time, and the original interaction log is converted into a computable numerical indicator to provide input for the calculation of function response efficiency. Based on the API call volume and transaction processing time, the function response efficiency is determined, the efficiency evaluation is simplified, and a side input is provided for the calculation of function dependency. Based on the function usage intensity and function response efficiency, the function dependency is determined to quantify the user's dependence on the platform function.

[0068] In some embodiments, multi-source market data is parsed to determine competitor data; competitor data is analyzed to determine the functional coverage of competitors; based on the functional coverage of competitors, the functional dependency impact is determined; based on the functional dependency impact, functional usage intensity, and functional response efficiency, the functional dependency degree is determined.

[0069] Competitive data can be dynamic information related to competitors.

[0070] Competitive feature coverage can be the breadth and depth of features provided by competitors.

[0071] Functional dependency impact can be the degree of negative impact that competitors have on user dependence on the platform's functions.

[0072] Specifically, a keyword matching algorithm is used to filter multi-source market data and extract competitor data. A feature extraction algorithm is then applied to the competitor data to identify the functional coverage of competing products. Subsequently, a platform-preset benchmark is established based on historical user data and preset reference thresholds based on industry experience. Furthermore, a gap analysis is conducted between the functional coverage of competing products and the platform-preset benchmark to determine the functional impact and dependency. Finally, a weighted integration algorithm is used to normalize the functional dependency impact, functional usage intensity, and functional response efficiency to determine the degree of functional dependency.

[0073] This solution parses multi-source market data, identifies competitor data, and ensures that external market factors are effectively isolated and prepared, avoiding the inability to fully analyze the interactions between data sources. Competitive data is analyzed to determine the coverage of competitor functions, ensuring that external competitive factors are included in the assessment and eliminating the problem of ignoring the dynamic constraints of competitors. Based on the coverage of competitor functions, the impact of functional dependencies is determined, ensuring that multi-source data is integrated to improve the accuracy of functional dependency quantification. Based on the impact of functional dependencies, functional usage intensity, and functional response efficiency, the degree of functional dependency is determined, eliminating the problem of being unable to fully analyze the interactions between data sources and providing reliable input for user stickiness assessment.

[0074] In some embodiments, the changing trend of functional dependency over time is analyzed to determine the functional dependency decay rate; user interaction data is parsed to determine user active cycle data; user behavior stability is determined based on user active cycle data; and user stickiness changes are determined based on the functional dependency decay rate and user behavior stability.

[0075] A change trend may be a pattern of changes in functional dependencies over time.

[0076] The functional dependency decay rate may be a trend of functional dependency decreasing over time.

[0077] User activity cycle data can be active period information parsed based on user interaction data.

[0078] User behavior stability can be a stability indicator calculated based on user activity cycle data.

[0079] Specifically, within the RSC platform's streaming computing engine, a time series analysis algorithm is applied to analyze functional dependencies. Furthermore, trend analysis is performed through QSC's IDAS, and the least squares method is used to fit a straight line to calculate the changing trend of functional dependencies over time. Subsequently, based on the analysis results, the functional dependency decay rate is determined. User interaction data is then cleaned and normalized through the streaming computing engine. Furthermore, user active periods are divided based on login frequency and API call volume, generating user activity cycle data. Volatility indicators are then calculated based on user activity cycle data, and user behavior stability is subsequently determined through normalization. Finally, the functional dependency decay rate and user behavior stability are weighted and integrated to determine changes in user stickiness.

[0080] This solution analyzes the changing trends of functional dependencies over time, determines the functional dependency decay rate, ensures the evolution of functional dependencies is quantified, and supports dynamic processing in the RSC platform's streaming computing engine, thereby enhancing the accuracy of user stickiness assessments. It also analyzes user interaction data to determine user activity cycle data, improving the basis for assessing user loyalty fluctuations and avoiding static biases. Based on user activity cycle data, it determines user behavior stability, ensures behavioral stability is quantified, and supports dynamic adjustments to user stickiness changes. Based on the functional dependency decay rate and user behavior stability, it determines user stickiness changes, ensuring that user stickiness change assessments integrate multi-source data in real time, thereby enhancing the timeliness of market decisions.

[0081] In some embodiments, multi-source market data is parsed to determine industry trend data; industry trend data is analyzed to determine a market volatility index; and user stickiness changes are determined based on the market volatility index, functional dependency decay rate, and user behavior stability.

[0082] Industry trend data can be structured market information parsed from multi-source market data.

[0083] The market volatility index can be a quantitative value that reflects the dynamic fluctuations of the market.

[0084] Specifically, the multi-source market data is formatted and standardized to extract fields such as industry growth rate, competitor feature coverage, and market share. Then, through time series alignment, filtering, and aggregation, industry trend data is generated. Subsequently, fluctuation detection based on historical data is applied to analyze dynamic changes in the industry trend data. Based on the analysis results, a market volatility index is quantified. Finally, the market volatility index, feature dependency decay rate, and user behavior stability are input into a linear weighted formula: the feature dependency decay rate is used as a negative factor, user behavior stability is used as a positive factor, and the market volatility index is used as a moderating factor to determine the final change in user stickiness.

[0085] This solution analyzes multi-source market data to identify industry trend data, avoiding the inability to fully analyze the interplay between data sources and ensuring that user stickiness change assessments are based on integrated multi-source data. By analyzing industry trend data and determining the market volatility index, we improve the accuracy and timeliness of user stickiness change predictions and support market decision-making. Based on the market volatility index, feature dependency decay rate, and user behavior stability, we determine user stickiness changes, improve the effectiveness of market analysis reports, and provide a basis for data-driven decision-making.

[0086] In some embodiments, the change in user stickiness is compared with a preset stickiness threshold to determine a low stickiness level where the user stickiness is lower than the preset stickiness threshold; multi-source market data corresponding to the low stickiness level is analyzed to determine the market positioning; based on the market positioning, the direction of user churn is determined; the difference in change between the low stickiness level and the preset stickiness threshold is analyzed; and based on the difference in change, the user churn risk level is determined.

[0087] The preset stickiness threshold may be a user stickiness threshold pre-stored in the RSC platform database, pre-stored in the server, and called when used.

[0088] User stickiness can be a quantitative indicator of user loyalty to the platform.

[0089] The low stickiness level may be a state where user stickiness is lower than a preset stickiness threshold.

[0090] Market positioning can be a quantitative position based on the functional coverage of competing products.

[0091] The change difference may be an absolute numerical difference between the stickiness low bit and a preset stickiness threshold.

[0092] Specifically, the preset stickiness threshold value set based on the statistical analysis of user stickiness in historical user churn data is accessed in the RSC database; then, the user stickiness change is compared with the preset stickiness threshold value in real time. If the user stickiness change is lower than the preset stickiness threshold value, it is marked as low stickiness. Subsequently, the multi-source market data associated with low stickiness is retrieved; then, the multi-source market data is parsed to extract the functional coverage of competing products; then, the market positioning is determined based on the functional coverage of competing products. Subsequently, the market positioning is mapped to the direction of user churn through the rule engine of IDAS. Then, the change difference between the low stickiness value and the preset stickiness threshold value is calculated. Finally, a preset rule is established based on the mapping relationship between the change difference value and the user churn risk level in historical churn cases; then, the change difference value is divided according to the preset rule to determine the user churn risk level.

[0093] This solution compares changes in user stickiness with preset stickiness thresholds, identifying low stickiness levels where user stickiness falls below the threshold. This quantifies whether user loyalty has reached a critical low point, ensuring that only cases with insufficient stickiness are further addressed, avoiding wasting resources on users with normal stickiness. Multi-source market data corresponding to low stickiness levels is analyzed to determine market positioning and reveal the external drivers of low stickiness, providing context for identifying the market root causes of declining user loyalty. Based on market positioning, the direction of user churn is determined, providing clear directional information on user churn, effectively focusing on targeted interventions, and thus improving the practicality of predictions. The difference in change between the low stickiness level and the preset stickiness threshold is analyzed to objectively measure the depth of insufficient stickiness and avoid subjective bias. Based on this difference in change, the user churn risk level is determined and the probability of user churn is quantified.

[0094] In some embodiments, the functional coverage of competing products is analyzed to determine the functional coverage between market positioning and competing products; user interaction data is parsed to determine transaction processing anomaly data; transaction processing anomaly data is analyzed to determine product functional defects; and the direction of user churn is determined based on functional coverage and product functional defects.

[0095] Competing products can be products or services from your competitors.

[0096] Functional coverage can be a quantified indicator of competitor functional coverage.

[0097] Transaction processing exception data may be exception records in user interaction data.

[0098] Product functional defects can be functional problems identified in transaction processing exception data.

[0099] Specifically, RSC's DMP is used to retrieve competitor product functional coverage from multi-source market data. The retrieved competitor product functional coverage is then parsed to extract indicators of competitor product functional coverage. Furthermore, the indicators of competitor product functional coverage are quantitatively compared with market positioning to calculate the functional coverage between market positioning and competitors. Subsequently, raw data is extracted from user interaction data. Pre-set anomaly detection rules are applied to the extracted raw data to determine transaction processing anomaly data. Pattern analysis is then performed on the transaction processing anomaly data, and a defect diagnosis algorithm is executed using RSC's IDAS to identify defect patterns in the transaction processing anomaly data. Product functional defects are then determined based on the defect patterns. Finally, preset mapping rules, established based on historical user churn data and empirical rules, are applied to map functional coverage and product functional defects to user churn directions.

[0100] This solution analyzes the functional coverage of competing products, determines the functional coverage of the platform compared to competitors in terms of market positioning, and identifies the platform's relative weaknesses in functional coverage. It also analyzes user interaction data to identify transaction anomalies, improving the real-time and accuracy of data processing. Transaction anomaly data analysis identifies product functional defects, providing root cause information for determining user churn. Based on functional coverage and product functional defects, user churn trends are determined, supporting churn risk prediction decisions.

[0101] In some embodiments, historical user churn data is obtained and analyzed to determine a churn risk baseline value; multi-source market data is analyzed to determine industry policy changes; based on industry policy changes, a policy impact coefficient is determined; and based on a weighted calculation of the change difference, the churn risk baseline value, and the policy impact coefficient, the user churn risk level is determined.

[0102] Historical user churn data may be stored record data of past user churn events.

[0103] The churn risk benchmark value may be a benchmark user churn probability value.

[0104] Industry policy changes can be industry-related policy adjustment events.

[0105] The policy impact coefficient may be a numerical coefficient that quantifies the degree to which changes in industry policies affect the risk of user churn.

[0106] Specifically, RSC's DMP was used to retrieve historical user churn data from the database. This data was then cleaned and standardized using RSC's IDAS. Statistical methods were then applied to analyze the historical churn data, extracting the proportion of churned users to the total number of users to determine the historical churn rate. The average of these historical churn rates was then used as the churn risk baseline. A pre-set rule engine based on natural language processing technology was then used to scan multi-source market data to identify policy-related keywords. Time series analysis was then used to compare current data with historical data to identify industry policy changes. Based on these industry policy changes, a pre-set rule library constructed through expert experience and validation of historical policy events was then applied, combined with the churn risk baseline to determine the policy impact coefficient. Finally, a multivariate linear regression was used, using historical user churn data as the dependent variable and the change difference, churn risk baseline, and policy impact coefficient as the independent variables to fit the regression coefficients. The change difference, churn risk baseline, and policy impact coefficient were then weighted and calculated, and the results were mapped to the user churn risk level.

[0107] This solution captures and analyzes historical user churn data to determine a churn risk baseline. This baseline serves as a quantitative benchmark for user churn risk prediction and is used as a benchmark input in weighted calculations. Multi-source market data is analyzed to identify industry policy changes, ensuring that external market factors are incorporated into risk analysis. Based on industry policy changes, a policy impact coefficient is determined to represent the positive or negative impact of the policy change on risk, which is used to adjust weights in weighted calculations. A weighted calculation of the change difference, the churn risk baseline, and the policy impact coefficient determines the user churn risk level, providing a quantitative indication of the probability of user churn and serving as a clear risk basis for market decision-making.

[0108] In some embodiments, transaction processing exception data is analyzed to determine the abnormal spatiotemporal distribution; based on the abnormal spatiotemporal distribution, the abnormal aggregation characteristics are determined; system architecture topology data is obtained, and based on the system architecture topology data, the abnormal association module is located; the abnormal aggregation characteristics are matched with the abnormal association module to obtain a matching relationship; based on the matching relationship, the defect propagation path is determined; based on the defect propagation path, the product function defect is determined.

[0109] The abnormal spatiotemporal distribution can be a structured graph representation of transaction processing abnormal data in the time dimension and space dimension.

[0110] The abnormal clustering feature can be a quantitative characteristic parameter of high-density clustering points in the abnormal spatiotemporal distribution.

[0111] System architecture topology data may be data describing the component hierarchical structure and connection relationships.

[0112] The anomaly association module may be a module that matches the anomaly's spatiotemporal distribution location.

[0113] The matching relationship may be a corresponding relationship between abnormal aggregation features and abnormal association modules.

[0114] The defect propagation path can be a dynamic path sequence of anomalies propagating between modules.

[0115] Specifically, a statistical analysis method is used to analyze transaction processing anomaly data and determine the spatiotemporal distribution of anomalies. First, in the temporal dimension, the frequency and distribution of anomalies are calculated by time window. Second, in the spatial dimension, spatial grids are divided according to location identifiers in the data, and the density of anomalies within each grid is calculated. Finally, the frequency and distribution are integrated with the density to form the spatiotemporal distribution of anomalies. Based on the spatiotemporal distribution of anomalies, a density detection method is applied to analyze the distribution map and identify anomaly clustering characteristics. Furthermore, system architecture topology data is obtained from the monitoring source through the RSC's DMP. This system architecture topology data is then parsed to create a module map. Based on the spatiotemporal distribution of anomalies, the locations of anomalies are matched with module nodes in the topology map to locate the anomaly-related modules. A rule matching engine is then used to correlate and match the anomaly clustering characteristics with the anomaly-related modules to form a matching relationship. Subsequently, the module sequence in the matching relationship is traversed, and causal inference methods are applied to determine the defect propagation path. Finally, the defect propagation path is parsed and the starting node in the defect propagation path is identified; then, through the mapping relationship between the starting node and the system function obtained from the system architecture topology data, the starting node is associated with the product function to determine the final product function defect.

[0116] This solution analyzes transaction processing anomaly data, determines the spatiotemporal distribution of anomalies, and effectively reveals the spatiotemporal hotspots where anomalies occur, avoiding one-sided analysis based on a single dimension. Based on the spatiotemporal distribution of anomalies, determine anomaly aggregation characteristics, and quantify the concentration and intensity characteristics of anomalies. Obtain system architecture topology data, locate anomaly-related modules based on the system architecture topology data, narrow the scope of the problem, and ensure that the focus is on the module where the anomaly actually occurs. Match the anomaly aggregation characteristics with the anomaly-related modules to obtain a matching relationship, and confirm evidence of anomalies in different modules under spatiotemporal conditions. Based on the matching relationship, determine the defect propagation path, reveal the dynamic propagation mechanism of the defect, and provide visual path data for identifying the source and diffusion point of the defect. Based on the defect propagation path, determine product functional defects, clarify the nature and impact of the defects, and improve the accuracy of user stickiness prediction.

[0117] Figure 3 A schematic diagram of the structure of a market management system based on RSC is provided in one embodiment of the present application. Figure 3 As shown, the RSC-based market management system 300 of this embodiment includes: a data acquisition module 301 , an interaction data analysis module 302 , a stickiness analysis module 303 , a churn analysis module 304 , and a report generation module 305 .

[0118] The data acquisition module 301 is used to acquire multi-source market data and user interaction data using RSC; Interaction data analysis module 302, used to analyze the user interaction data and determine functional dependencies; A stickiness analysis module 303 is used to determine changes in user stickiness based on the functional dependency; A churn analysis module 304 is configured to analyze the multi-source market data based on the changes in user stickiness and predict the user churn risk level and user churn direction; The report generating module 305 is configured to generate a market analysis report according to the user churn risk level and the user churn direction.

[0119] Optionally, when analyzing the user interaction data to determine the functional dependency, the interaction data analysis module 302 is configured to: Parsing the user interaction data to determine user operation data and response interaction data; Analyze the user operation data to determine the login frequency and function usage depth; Determining the intensity of function usage based on the login frequency and the depth of function usage; Analyze the response interaction data to determine the API call volume and transaction processing time; Determine the function response efficiency based on the API call volume and the transaction processing time; The function dependency is determined according to the function usage intensity and the function response efficiency.

[0120] Optionally, when determining the function dependency based on the function usage intensity and the function response efficiency, the interaction data analysis module 302 is configured to: Analyze the multi-source market data to determine competing product data; Analyze the competitor data to determine the functional coverage of the competitor; Determine the impact of functional dependencies based on the functional coverage of the competitor product; The function dependency is determined according to the function dependency impact, the function usage intensity and the function response efficiency.

[0121] Optionally, when determining a change in user stickiness based on the functional dependency, the stickiness analysis module 303 is configured to: Analyzing a changing trend of the functional dependency over time to determine a functional dependency decay rate; Analyzing the user interaction data to determine user activity period data; Determining user behavior stability based on the user activity cycle data; The user stickiness change is determined based on the function dependency decay rate and the user behavior stability.

[0122] Optionally, when determining a change in user stickiness based on the function dependency decay rate and the user behavior stability, the stickiness analysis module 303 is configured to: Analyze the multi-source market data to determine industry trend data; Analyze the industry trend data to determine the market volatility index; The change in user stickiness is determined based on the market volatility index, the function dependency decay rate, and the user behavior stability.

[0123] Optionally, when the churn analysis module 304 analyzes the multi-source market data based on the change in user stickiness and predicts the user churn risk level and user churn direction, it is configured to: Comparing the user stickiness change with a preset stickiness threshold, and determining a low stickiness level where the user stickiness is lower than the preset stickiness threshold; Analyze multi-source market data corresponding to the aforementioned low stickiness to determine market positioning; Determine the direction of user churn based on the market positioning; Analyze the change difference between the low stickiness level and a preset stickiness threshold; and determine the user churn risk level based on the change difference.

[0124] Optionally, when determining the user churn direction based on the market positioning, the churn analysis module 304 is configured to: Analyze the functional coverage of the competing products and determine the functional coverage between the market positioning and the competing products; parsing the user interaction data to determine transaction processing exception data; Analyze the transaction processing exception data to determine product functional defects; Determine the direction of user loss based on the functional coverage and product functional defects.

[0125] Optionally, when determining the user churn risk level based on the change difference, the churn analysis module 304 is configured to: Obtaining historical user churn data, analyzing the historical user churn data, and determining a churn risk baseline value; Analyze the multi-source market data to identify changes in industry policies; Determine the policy impact coefficient based on the industry policy changes; The user churn risk level is determined based on a weighted calculation of the change difference, the churn risk baseline value, and the policy impact coefficient.

[0126] Optionally, when the churn analysis module 304 analyzes the transaction processing exception data and determines a product function defect, it is used to: Analyzing the transaction processing exception data to determine the abnormal temporal and spatial distribution; Determining abnormal aggregation characteristics based on the abnormal spatiotemporal distribution; Acquire system architecture topology data, and locate abnormal related modules based on the system architecture topology data; Matching the abnormal aggregation feature with the abnormal association module to obtain a matching relationship; determining a defect propagation path according to the matching relationship; Determine product functional defects based on the defect propagation path.

[0127] The system of this embodiment can be used to execute the method of any of the above embodiments. Its implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A market management method based on RSC, characterized in that: include: Use RSC to obtain multi-source market data and user interaction data; Analyzing the user interaction data to determine functional dependencies; Determining a change in user stickiness based on the functional dependency; Analyze the multi-source market data based on the changes in user stickiness to predict the user churn risk level and user churn direction; Generate a market analysis report based on the user churn risk level and the user churn direction.

2. The method according to claim 1, characterized in that The analyzing the user interaction data to determine the functional dependency includes: Parsing the user interaction data to determine user operation data and response interaction data; Analyze the user operation data to determine the login frequency and function usage depth; Determining the intensity of function usage based on the login frequency and the depth of function usage; Analyze the response interaction data to determine the API call volume and transaction processing time; Determine the function response efficiency based on the API call volume and the transaction processing time; The function dependency is determined according to the function usage intensity and the function response efficiency.

3. The method according to claim 2, characterized in that The determining of the function dependency according to the function usage intensity and the function response efficiency includes: Analyze the multi-source market data to determine competing product data; Analyze the competitor data to determine the functional coverage of the competitor; Determine the impact of functional dependencies based on the functional coverage of the competitor product; The function dependency is determined according to the function dependency impact, the function usage intensity and the function response efficiency.

4. The method according to claim 1, wherein The determining of the user stickiness change according to the functional dependency includes: Analyzing a changing trend of the functional dependency over time to determine a functional dependency decay rate; Analyzing the user interaction data to determine user activity period data; Determining user behavior stability based on the user activity cycle data; The user stickiness change is determined based on the function dependency decay rate and the user behavior stability.

5. The method according to claim 4, characterized in that The determining of the user stickiness change according to the function dependency decay rate and the user behavior stability includes: Analyze the multi-source market data to determine industry trend data; Analyze the industry trend data to determine the market volatility index; The change in user stickiness is determined based on the market volatility index, the function dependency decay rate, and the user behavior stability.

6. The method according to claim 3, characterized in that The step of analyzing the multi-source market data based on the changes in user stickiness to predict the user churn risk level and user churn direction includes: Comparing the user stickiness change with a preset stickiness threshold, and determining a low stickiness level where the user stickiness is lower than the preset stickiness threshold; Analyze multi-source market data corresponding to the aforementioned low stickiness to determine market positioning; Determine the direction of user churn based on the market positioning; Analyze the change difference between the low stickiness level and a preset stickiness threshold; and determine the user churn risk level based on the change difference.

7. The method according to claim 6, characterized in that Determining the direction of user churn based on the market positioning includes: Analyze the functional coverage of the competing products and determine the functional coverage between the market positioning and the competing products; parsing the user interaction data to determine transaction processing exception data; Analyze the transaction processing exception data to determine product functional defects; Determine the direction of user loss based on the functional coverage and product functional defects.

8. The method according to claim 6, characterized in that Determining the user churn risk level according to the change difference includes: Obtaining historical user churn data, analyzing the historical user churn data, and determining a churn risk baseline value; Analyze the multi-source market data to identify changes in industry policies; Determine the policy impact coefficient based on the industry policy changes; The user churn risk level is determined based on a weighted calculation of the change difference, the churn risk baseline value, and the policy impact coefficient.

9. The method according to claim 7, characterized in that The analyzing the transaction processing exception data to determine product functional defects includes: Analyzing the transaction processing exception data to determine the abnormal temporal and spatial distribution; Determining abnormal aggregation characteristics based on the abnormal spatiotemporal distribution; Acquire system architecture topology data, and locate abnormal related modules based on the system architecture topology data; Matching the abnormal aggregation feature with the abnormal association module to obtain a matching relationship; determining a defect propagation path according to the matching relationship; Determine product functional defects based on the defect propagation path.

10. A market management system based on RSC, characterized in that: The method as claimed in any one of claims 1 to 9 comprises: Data acquisition module, used to obtain multi-source market data and user interaction data using RSC; An interaction data analysis module, configured to analyze the user interaction data and determine functional dependencies; A stickiness analysis module, configured to determine changes in user stickiness based on the functional dependency; A churn analysis module, configured to analyze the multi-source market data based on the changes in user stickiness, and predict the user churn risk level and user churn direction; The report generation module is used to generate a market analysis report based on the user churn risk level and the user churn direction.

Images