Mobile application platform optimization decision-making method and system based on evaluation results

Abstract

The application discloses a mobile application platform optimization decision-making method and system based on evaluation results, and belongs to the technical field of mobile application quality management and control and intelligent optimization.The application constructs a multidimensional comprehensive evaluation system covering user experience, business energy efficiency, running performance, stability, maturity and activity, adopts a non-invasive probe to complete full-quantity data collection, realizes quantitative evaluation through a weighted comprehensive scoring algorithm, completes accurate positioning of abnormal root causes based on a coupling algorithm of an improved analytic hierarchy process-entropy weight method, automatically matches an optimization strategy library to generate a hierarchical optimization decision-making scheme, and forms a full-closed-loop management of evaluation-analysis-optimization-verification.The application solves the problems of fragmentation of the evaluation system, low accuracy of root cause positioning and lack of a closed-loop optimization mechanism in the prior art, is suitable for enterprise-level mobile application platform whole-life-cycle quality management and control, and can significantly improve application stability, operation and maintenance efficiency and user satisfaction.

Description

Technical Field

[0001] This invention belongs to the field of mobile application quality control and intelligent optimization technology, specifically relating to a mobile application platform optimization decision-making method and system based on evaluation results. Background Technology

[0002] With the deepening of enterprise digital transformation, mobile application platforms have become the core carrier for business operations and user services in industries such as power. The quality, user experience, and stability of mobile applications directly affect business efficiency and user satisfaction. Current technologies for managing mobile application platforms mostly focus on basic operational monitoring and anomaly alerts, lacking a systematic and comprehensive evaluation mechanism adapted to industry characteristics, and also exhibiting the following technical shortcomings:

[0003] 1. The evaluation system is fragmented. Existing evaluations often focus solely on performance or stability, failing to integrate multi-dimensional indicators such as user experience, business efficiency, maturity, and activity, thus failing to comprehensively reflect the overall quality of mobile applications. In particular, for mobile applications in the power industry, differentiated indicator weights and evaluation benchmarks have not been set according to the characteristics of business scenarios, resulting in evaluation results that are out of touch with actual business practices.

[0004] 2. Insufficient data collection and governance capabilities. Existing data collection relies heavily on manual data entry on the business side, which is highly intrusive, costly to modify, and cannot achieve standardized processing of multi-source heterogeneous data, making it difficult to support accurate quantitative evaluation.

[0005] 3. The lack of a closed-loop decision-making mechanism from evaluation to optimization means that existing technologies can only detect and alert on problems, but cannot automatically locate the root cause based on the evaluation results, let alone generate targeted optimization decision-making solutions. They rely on the manual experience analysis of technical personnel, resulting in low optimization efficiency, delayed handling, and inability to achieve continuous iterative improvement of mobile application quality.

[0006] 4. The root cause analysis is not precise enough. Existing anomaly localization methods mostly use single rule matching, which cannot quantify the contribution of each influencing factor to the anomaly index, making it difficult to locate the root cause of the deep problem. This results in poor targeting of optimization solutions and the inability to guarantee optimization results.

[0007] In view of the above-mentioned defects in the existing technology, the present invention aims to provide a mobile application platform optimization decision-making method and system based on evaluation results, which solves the technical problems of incomplete evaluation system, low root cause location accuracy and lack of closed-loop optimization decision-making mechanism in the existing technology, and realizes the continuous improvement of mobile application platform quality and experience. Summary of the Invention

[0008] The purpose of this invention is to provide a mobile application platform optimization decision-making method and system based on evaluation results, so as to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a mobile application platform optimization decision-making method based on evaluation results, comprising the following steps: S1. Construct a multi-dimensional comprehensive evaluation index system for mobile applications, which includes user experience dimension, business efficiency dimension, operating performance dimension, stability dimension, maturity and activity dimension; S2. Collect all operational and business data of the mobile application platform for each evaluation dimension through non-intrusive data probes, and complete data cleaning and standardization processing. S3. Based on the standardized indicator data, calculate the sub-scores for each dimension and the comprehensive evaluation score of the mobile application, and generate multi-dimensional visual evaluation results. S4. Based on the comprehensive evaluation results and sub-item scores, identify abnormal indicators that are below the preset threshold, and use the root cause analysis algorithm to locate the core influencing factors and root causes of the abnormal indicators. S5. Based on the root cause of the problem, match the preset optimization strategy library, generate hierarchical optimization decision schemes, and push them to the corresponding execution entities; S6. Track the implementation effect of the optimized decision-making scheme, complete the evaluation loop based on the optimized indicator data, and dynamically update the evaluation system and optimization strategy library.

[0010] It should be noted in the solution that the weighted comprehensive scoring algorithm formula used in step S3 to calculate the comprehensive evaluation score of the mobile application is as follows: ; in, This is the overall evaluation score for mobile applications, with a value range of 0-100. The total number of evaluation dimensions. These correspond to the user experience dimension, business efficiency dimension, operational performance dimension, stability dimension, and maturity and activity dimension, respectively. For the first The weight coefficients of each evaluation dimension satisfy the following: Furthermore, the weight coefficients for each dimension can be customized based on industry characteristics and business scenarios; For the first The standardized scores for each evaluation dimension range from 0 to 100, with individual dimension scores... The calculation formula is: ; In the formula, For the first The number of third-level indicators under each evaluation dimension For the first The first dimension The weighting coefficients of the three-level indicators satisfy the following: ; For the first The first dimension The standardized values ​​of the three-level indicators are calculated using the min-max standardization method. The standardization formula for positive indicators (the larger the value, the better the performance) is as follows: ; The standardized formula for negative indicators (smaller values ​​indicate better performance) is: ; in, For the first The actual collected values ​​of each of the three-level indicators. , The first The maximum and minimum values ​​of each tertiary indicator within the statistical period.

[0011] It is further worth noting that in step S4, the core influencing factors are located using a root cause analysis algorithm. The root cause localization algorithm is a modified analytic hierarchy process coupled with entropy weighting. The specific formula and calculation steps are as follows: Step 1: Construct the influence factor judgment matrix of abnormal indicators ,in This represents the number of candidate influencing factors corresponding to the abnormal indicators. Impact Factor Relative to impact factor The importance scale is assigned using a 1-9 scale. Step 2: Calculate the subjective weights of each influencing factor using the analytic hierarchy process (AHP). The consistency of the judgment matrix is ​​checked, and the judgment matrix is ​​valid when the consistency ratio CR < 0.1. Step 3: Calculate the objective weight of each influencing factor using the entropy weight method. The formula is as follows: ; in, , Impact Factor and abnormal indicators The Pearson correlation coefficient, if Then define ; ; In the formula, Impact Factor Information entropy Impact Factor Objective entropy weight; Step 4: Combine subjective and objective weights to calculate the combined weight of the influencing factors. The formula is: ; Step 5: By combination weight Sort the factors from largest to smallest, and select the top N factors as the core influencing factors of the abnormal indicators to complete the root cause location of the problem.

[0012] Furthermore, it should be noted that the multi-dimensional comprehensive evaluation index system constructed in step S1 includes user experience dimensions such as ease of operation, page smoothness, APDEX experience, scope of impact of lag, scope of impact of crash, and user-initiated evaluation index. Business efficiency metrics include active user count, user usage time, business process pass rate, business conversion rate, and version update quality indicators. Performance metrics include page load time, network request time, application startup time, terminal power consumption, and method execution time. Stability metrics include crash rate, ANR rate, JS error rate, page loading error rate, and operation error rate. The maturity and activity dimensions include application launch frequency, session activity, feature coverage, and iteration optimization frequency.

[0013] As a preferred implementation, the non-intrusive data probe in step S2 supports mainstream mobile operating systems such as Android, iOS, and HarmonyOS. It achieves no-code data collection through SDK integration. The collected content includes user behavior data, page rendering data, network request data, application crash logs, system operation data, and business process data. The collection process does not modify the native code of the business application.

[0014] As a preferred implementation, the hierarchical optimization decision scheme generated in step S5 is divided into three levels: emergency optimization, routine optimization, and iterative optimization.

[0015] Emergency optimizations are for scenarios with an overall score below 60 or core stability indicators exceeding the threshold, and are pushed to the application operations and development team for immediate handling; regular optimizations are for scenarios with an overall score of 60-80, and are included in the monthly optimization plan; iterative optimizations are for experience improvement indicators with an overall score of 80 or above, and are included in the application version iteration plan.

[0016] As a preferred implementation, the evaluation loop in step S6 specifically includes:

[0017] Compare the corresponding indicator data before and after the optimization decision is executed, calculate the optimization efficiency, and when the optimization efficiency is lower than the preset threshold, re-execute the root cause analysis and optimization scheme generation.

[0018] At the same time, the effective optimization schemes will be added to the optimization strategy library, and the weight coefficients of each indicator in the evaluation system will be dynamically adjusted based on the optimized indicator distribution.

[0019] This invention also provides the following technical solution: a mobile application platform optimization decision-making system based on evaluation results, used to implement the mobile application platform optimization decision-making method based on evaluation results described in any one of the above claims, comprising: The indicator system construction module is used to construct a multi-dimensional comprehensive evaluation indicator system for mobile applications. The indicator system includes user experience dimension, business efficiency dimension, operational performance dimension, stability dimension, maturity and activity dimension. The data acquisition and processing module is used to collect full operational and business data of the mobile application platform for each evaluation dimension through non-intrusive data probes, and to complete data cleaning and standardization processing. The comprehensive evaluation module is used to calculate the sub-scores for each dimension and the comprehensive evaluation score of the mobile application based on the standardized indicator data, and generate multi-dimensional visual evaluation results. The root cause localization module is used to identify abnormal indicators below a preset threshold based on the comprehensive evaluation results and sub-item scores, and to locate the core influencing factors and root causes of the problems corresponding to the abnormal indicators through the root cause analysis algorithm. The optimization decision generation module is used to match the root cause of the problem with a preset optimization strategy library, generate hierarchical optimization decision schemes, and push them to the corresponding execution entities. The closed-loop verification and iteration module is used to track the execution effect of the optimized decision-making scheme, complete the evaluation closed loop based on the optimized indicator data, and dynamically update the evaluation system and optimization strategy library.

[0020] In a preferred embodiment, the comprehensive evaluation module has a built-in weighted comprehensive scoring algorithm unit, which executes the algorithm formula described in step S3 to calculate the comprehensive evaluation score; the root cause localization module has a built-in coupled root cause analysis algorithm unit, which executes the algorithm formula described in step S4 to locate the core influencing factors.

[0021] Compared with the prior art, the mobile application platform optimization decision-making method and system based on evaluation results provided by the present invention have at least the following beneficial effects:

[0022] (1) This invention constructs a comprehensive evaluation system that integrates five dimensions: user experience, business efficiency, operational performance, stability, maturity and activity. It covers the quality control nodes of the entire life cycle of mobile applications. At the same time, it can customize the indicator weights and evaluation benchmarks based on the business characteristics of industries such as power. The evaluation results are comprehensive and in line with the actual business, filling the gap in the industry-level mobile application quality quantitative evaluation standards. Furthermore, it adopts non-intrusive data probes to achieve no-code data collection, supports mainstream mobile operating systems, and does not require modification of the native code of business applications, achieving seamless integration of business applications and significantly reducing the transformation cost of data collection. At the same time, through standardized data processing, it solves the problem of difficult governance of multi-source heterogeneous data, providing reliable data support for accurate evaluation.

[0023] (2) This invention designs a weighted comprehensive scoring algorithm coupled with an improved hierarchical analysis-entropy weight method to locate the root cause. It achieves accurate quantification of the comprehensive score by combining subjective and objective methods. At the same time, it can accurately locate the core influencing factors of abnormal indicators. The root cause location accuracy is high, which solves the problem of existing technologies relying on human experience and inaccurate location. Furthermore, it establishes a closed-loop decision-making mechanism from evaluation to optimization. Based on the root cause location results, it automatically matches the optimization strategy library, generates hierarchical optimization decision schemes, clarifies the priority of disposal and the execution subject, and tracks the optimization effect. It dynamically iterates the evaluation system and optimization strategy library, realizing the automated and intelligent continuous optimization of mobile application quality, and greatly improving operation and maintenance efficiency and user satisfaction.

[0024] (3) The system of the present invention has high scalability and high availability, can be adapted to enterprise-level mobile application platforms with 500,000 registered users, meet the high concurrency access requirements, and is compatible with the enterprise's existing unified authentication system, cloud services and data center. It has strong adaptability, meets the requirements of network security level protection and industry digital construction standards, and can be widely used in the management and control of enterprise-level mobile application platforms in the power and other industries. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the steps of the mobile application platform optimization decision-making method based on evaluation results as described in this invention.

[0026] Figure 2 This is a module architecture diagram of the mobile application platform optimization decision system based on evaluation results described in this invention. Detailed Implementation

[0027] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0028] Please see Figure 1This invention provides a mobile application platform optimization decision-making method based on evaluation results, comprising the following steps:

[0029] S1. Construct a multi-dimensional comprehensive evaluation index system for mobile applications. The index system includes user experience dimension, business efficiency dimension, operating performance dimension, stability dimension, maturity and activity dimension. Among them, the user experience dimension includes ease of operation, page smoothness, APDEX experience, lag impact range, crash impact range, and user-initiated evaluation index. Business efficiency metrics include active user count, user usage time, business process pass rate, business conversion rate, and version update quality indicators. Performance metrics include page load time, network request time, application startup time, terminal power consumption, and method execution time. Stability metrics include crash rate, ANR rate, JS error rate, page loading error rate, and operation error rate. The maturity and activity dimensions include application launch frequency, session activity, feature coverage, and iteration optimization frequency.

[0030] S2. Collect full operational and business data of the mobile application platform for each evaluation dimension through non-intrusive data probes, and complete data cleaning and standardization processing.

[0031] Among them, the non-intrusive data probe supports mainstream mobile operating systems such as Android, iOS, and HarmonyOS. It achieves no-code data collection through SDK integration. The collected content includes user behavior data, page rendering data, network request data, application crash logs, system operation data, and business process data. The collection process does not modify the native code of the business application.

[0032] S3. Based on the standardized indicator data, calculate the sub-scores for each dimension and the comprehensive evaluation score of the mobile application, and generate multi-dimensional visual evaluation results. Specifically, the weighted comprehensive scoring algorithm formula used to calculate the overall evaluation score of mobile applications is as follows: ; in, This is the overall evaluation score for mobile applications, with a value range of 0-100. The total number of evaluation dimensions. These correspond to the user experience dimension, business efficiency dimension, operational performance dimension, stability dimension, and maturity and activity dimension, respectively. For the first The weight coefficients of each evaluation dimension satisfy the following: Furthermore, the weight coefficients for each dimension can be customized based on industry characteristics and business scenarios; For the first The standardized scores for each evaluation dimension range from 0 to 100, with individual dimension scores... The calculation formula is: ; In the formula, For the first The number of third-level indicators under each evaluation dimension For the first The first dimension The weighting coefficients of the three-level indicators satisfy the following: ; For the first The first dimension The standardized values ​​of the three-level indicators are calculated using the min-max standardization method. The standardization formula for positive indicators (the larger the value, the better the performance) is as follows: ; The standardized formula for negative indicators (smaller values ​​indicate better performance) is: ; in, For the first The actual collected values ​​of each of the three-level indicators. , The first The maximum and minimum values ​​of each tertiary indicator within the statistical period.

[0033] S4. Based on the comprehensive evaluation results and sub-item scores, identify abnormal indicators that are below the preset threshold, and use the root cause analysis algorithm to locate the core influencing factors and root causes of the abnormal indicators. Specifically, the root cause analysis algorithm is used to locate the core influencing factors. An improved analytic hierarchy process (AHP) coupled with entropy weighting is employed as the root cause localization algorithm. The specific formula and calculation steps are as follows: Step 1: Construct the influence factor judgment matrix of abnormal indicators ,in This represents the number of candidate influencing factors corresponding to the abnormal indicators. Impact Factor Relative to impact factor The importance scale is assigned using a 1-9 scale. Step 2: Calculate the subjective weights of each influencing factor using the analytic hierarchy process (AHP). The consistency of the judgment matrix is ​​checked, and the judgment matrix is ​​valid when the consistency ratio CR < 0.1. Step 3: Calculate the objective weight of each influencing factor using the entropy weight method. The formula is as follows: ; in, , Impact Factor and abnormal indicators The Pearson correlation coefficient, if Then define ; ; In the formula, Impact Factor Information entropy Impact Factor Objective entropy weight; Step 4: Combine subjective and objective weights to calculate the combined weight of the influencing factors. The formula is: ; Step 5: By combination weight Sort the factors from largest to smallest, and select the top N factors as the core influencing factors of the abnormal indicators to complete the root cause location of the problem.

[0034] S5. Based on the root cause of the problem, match the preset optimization strategy library, generate hierarchical optimization decision schemes, and push them to the corresponding execution entities.

[0035] Specifically, the hierarchical optimization decision-making scheme is divided into three levels: emergency optimization, routine optimization, and iterative optimization.

[0036] Emergency optimizations are for scenarios with an overall score below 60 or core stability indicators exceeding the threshold, and are pushed to the application operations and development team for immediate handling; regular optimizations are for scenarios with an overall score of 60-80, and are included in the monthly optimization plan; iterative optimizations are for experience improvement indicators with an overall score of 80 or above, and are included in the application version iteration plan.

[0037] S6. Track the implementation effect of the optimized decision-making scheme, complete the evaluation loop based on the optimized indicator data, and dynamically update the evaluation system and optimization strategy library.

[0038] The evaluation closed loop is specifically as follows:

[0039] Compare the corresponding indicator data before and after the optimization decision is executed, calculate the optimization efficiency, and when the optimization efficiency is lower than the preset threshold, re-execute the root cause analysis and optimization scheme generation.

[0040] At the same time, the effective optimization schemes will be added to the optimization strategy library, and the weight coefficients of each indicator in the evaluation system will be dynamically adjusted based on the optimized indicator distribution.

[0041] Please see Figure 2This invention provides a mobile application platform optimization decision-making system based on evaluation results, including an indicator system construction module, a data acquisition and processing module, a comprehensive evaluation module, a root cause localization module, an optimization decision generation module, and a closed-loop verification and iteration module. The specific implementation methods of each module are as follows:

[0042] Indicator System Construction Module: Used to build a multi-dimensional comprehensive evaluation indicator system for mobile applications. It supports custom dimensions, indicators, weight coefficients, evaluation thresholds, and industry benchmark values. It has a built-in standard evaluation template for mobile applications in the power industry and supports configuring differentiated evaluation systems according to application type and business scenario. It complies with the relevant technical specifications for digital construction of China Southern Power Grid Company.

[0043] Data Acquisition and Processing Module: Includes a non-intrusive data probe unit, a data cleaning unit, and a data standardization unit. The data probe unit supports no-code data acquisition on mainstream operating systems such as Android, iOS, and HarmonyOS, and supports probe enable / disable management, version management, and deployment management. The data cleaning unit is used to complete the deduplication, invalid value removal, and outlier filtering of raw data. The data standardization unit has a built-in min-max standardization algorithm to complete the standardization processing of multi-source heterogeneous data.

[0044] Comprehensive Evaluation Module: Includes a built-in weighted comprehensive scoring algorithm unit and a visualization display unit; the weighted comprehensive scoring algorithm unit executes the algorithm formula described in step S3 to automatically calculate the sub-item scores and comprehensive evaluation scores for each dimension; the visualization display unit is used to generate multi-dimensional evaluation dashboards, trend charts, and detailed reports, supporting drill-down queries by application, unit, and time dimension, while also supporting custom report settings and periodic automatic push.

[0045] Root cause localization module: It has a built-in abnormal indicator identification unit and a coupled root cause analysis algorithm unit. The abnormal indicator identification unit is used to compare the indicator score with the preset threshold and automatically identify abnormal indicators. The coupled root cause analysis algorithm unit executes the algorithm formula described in step S4 to complete the calculation and sorting of the core influencing factors of abnormal indicators, locate the root cause of the problem, and output a root cause analysis report.

[0046] The optimization decision generation module includes a built-in optimization strategy library, a hierarchical decision unit, and a push unit. The optimization strategy library stores standardized optimization strategies, handling procedures, and verification standards for each abnormal indicator, and supports adding, editing, searching, and iteratively updating strategies. The hierarchical decision unit generates optimization decision plans at three levels—emergency optimization, regular optimization, and iterative optimization—based on the abnormality level and comprehensive score. The push unit pushes the optimization decision plans to the corresponding implementing entities, supporting various push methods such as system notifications and emails.

[0047] The closed-loop verification and iteration module includes a built-in effect tracking unit, a strategy iteration unit, and an evaluation system iteration unit. The effect tracking unit collects optimized indicator data, compares the changes in indicators before and after optimization, calculates the optimization effectiveness, and completes the optimization effect verification. The strategy iteration unit adds verified and effective optimization schemes to the optimization strategy library to improve the strategy content. The evaluation system iteration unit dynamically adjusts the indicator weights and benchmark values ​​of the evaluation system based on the distribution of all indicator data to achieve continuous optimization of the evaluation system.

[0048] This system meets the following performance and security requirements: it supports no fewer than 50,000 concurrent online users, with no fewer than 500 concurrent user logins and single sign-on, a normal page response time of no more than 3 seconds, and an annual system availability of no less than 99.9%; it has horizontal scalability and can be adapted to the enterprise's existing unified authentication system, Southern Power Grid Cloud Platform, data center, and other basic environments; it meets the requirements of Level 2 Information Security Protection and complies with relevant laws and standards such as the Cybersecurity Law and the Administrative Measures for Cybersecurity Protection in the Power Industry.

[0049] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0050] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0051] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A mobile application platform optimization decision-making method based on evaluation results, characterized in that: Includes the following steps: S1. Construct a multi-dimensional comprehensive evaluation index system for mobile applications, which includes user experience dimension, business efficiency dimension, operating performance dimension, stability dimension, maturity and activity dimension; S2. Collect all operational and business data of the mobile application platform for each evaluation dimension through non-intrusive data probes, and complete data cleaning and standardization processing. S3. Based on the standardized indicator data, calculate the sub-scores for each dimension and the comprehensive evaluation score of the mobile application, and generate multi-dimensional visual evaluation results. S4. Based on the comprehensive evaluation results and sub-item scores, identify abnormal indicators that are below the preset threshold, and use the root cause analysis algorithm to locate the core influencing factors and root causes of the abnormal indicators. S5. Based on the root cause of the problem, match the preset optimization strategy library, generate hierarchical optimization decision schemes, and push them to the corresponding execution entities; S6. Track the implementation effect of the optimized decision-making scheme, complete the evaluation loop based on the optimized indicator data, and dynamically update the evaluation system and optimization strategy library.

2. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: In step S3, the weighted comprehensive scoring algorithm used to calculate the mobile application's overall evaluation score is as follows: ； in, This is the overall evaluation score for mobile applications, with a value range of 0-100. The total number of evaluation dimensions. These correspond to the user experience dimension, business efficiency dimension, operational performance dimension, stability dimension, and maturity and activity dimension, respectively. For the first The weight coefficients of each evaluation dimension satisfy the following: Furthermore, the weight coefficients for each dimension can be customized based on industry characteristics and business scenarios; For the first The standardized scores for each evaluation dimension range from 0 to 100, with individual dimension scores... The calculation formula is: ； In the formula, For the first The number of third-level indicators under each evaluation dimension For the first The first dimension The weighting coefficients of the three-level indicators satisfy the following: ; For the first The first dimension The standardized values ​​of the three-level indicators are calculated using the min-max standardization method. The standardization formula for positive indicators (the larger the value, the better the performance) is as follows: ； The standardized formula for negative indicators (smaller values ​​indicate better performance) is: ； in, For the first The actual collected values ​​of each of the three-level indicators. , The first The maximum and minimum values ​​of each tertiary indicator within the statistical period.

3. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: In step S4, the core influencing factors are located using a root cause analysis algorithm. The improved analytic hierarchy process (AHP) coupled with entropy weighting is employed as the root cause localization algorithm. The specific formula and calculation steps are as follows: Step 1: Construct the influence factor judgment matrix of abnormal indicators ,in This represents the number of candidate influencing factors corresponding to the abnormal indicators. Impact Factor Relative to impact factor The importance scale is assigned using a 1-9 scale. Step 2: Calculate the subjective weights of each influencing factor using the analytic hierarchy process (AHP). The consistency of the judgment matrix is ​​checked, and the judgment matrix is ​​valid when the consistency ratio CR < 0.

1. Step 3: Calculate the objective weight of each influencing factor using the entropy weight method. The formula is as follows: ； in, , Impact Factor and abnormal indicators The Pearson correlation coefficient, if Then define ; ； In the formula, Impact Factor Information entropy Impact Factor Objective entropy weight; Step 4: Combine subjective and objective weights to calculate the combined weight of the influencing factors. The formula is: ； Step 5: By combination weight Sort the factors from largest to smallest, and select the top N factors as the core influencing factors of the abnormal indicators to complete the root cause location of the problem.

4. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: The multi-dimensional comprehensive evaluation index system constructed in step S1 includes user experience dimensions such as ease of operation, page smoothness, APDEX experience, scope of impact of lag, scope of impact of crash, and user-initiated evaluation index. Business efficiency metrics include active user count, user usage time, business process pass rate, business conversion rate, and version update quality indicators. Performance metrics include page load time, network request time, application startup time, terminal power consumption, and method execution time. Stability metrics include crash rate, ANR rate, JS error rate, page loading error rate, and operation error rate. The maturity and activity dimensions include application launch frequency, session activity, feature coverage, and iteration optimization frequency.

5. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: The non-intrusive data probe in step S2 supports mainstream mobile operating systems such as Android, iOS, and HarmonyOS. It achieves no-code data collection through SDK integration. The collected content includes user behavior data, page rendering data, network request data, application crash logs, system operation data, and business process data. The collection process does not modify the native code of the business application.

6. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: The hierarchical optimization decision scheme generated in step S5 is divided into three levels: emergency optimization, routine optimization, and iterative optimization. Emergency optimizations are for scenarios where the overall score is below 60 or the core stability indicators exceed the threshold, and are pushed to the application operation and development team for immediate handling. Regular optimizations are applied to scenarios with a comprehensive score of 60-80 points and are included in the monthly optimization plan. Iterative optimizations that correspond to experience improvement metrics with a comprehensive score of 80 or above will be incorporated into the application version iteration plan.

7. The mobile application platform optimization decision-making method based on evaluation results according to claim 1, characterized in that: The evaluation loop in step S6 is specifically as follows: Compare the corresponding indicator data before and after the optimization decision is executed, calculate the optimization efficiency, and when the optimization efficiency is lower than the preset threshold, re-execute the root cause analysis and optimization scheme generation. At the same time, the effective optimization schemes will be added to the optimization strategy library, and the weight coefficients of each indicator in the evaluation system will be dynamically adjusted based on the optimized indicator distribution.

8. A mobile application platform optimization decision-making system based on evaluation results, used to implement the mobile application platform optimization decision-making method based on evaluation results as described in any one of claims 1-7, characterized in that: include: The indicator system construction module is used to construct a multi-dimensional comprehensive evaluation indicator system for mobile applications. The indicator system includes user experience dimension, business efficiency dimension, operational performance dimension, stability dimension, maturity and activity dimension. The data acquisition and processing module is used to collect full operational and business data of the mobile application platform for each evaluation dimension through non-intrusive data probes, and to complete data cleaning and standardization processing. The comprehensive evaluation module is used to calculate the sub-scores for each dimension and the comprehensive evaluation score of the mobile application based on the standardized indicator data, and generate multi-dimensional visual evaluation results. The root cause localization module is used to identify abnormal indicators below a preset threshold based on the comprehensive evaluation results and sub-item scores, and to locate the core influencing factors and root causes of the problems corresponding to the abnormal indicators through the root cause analysis algorithm. The optimization decision generation module is used to match the root cause of the problem with a preset optimization strategy library, generate hierarchical optimization decision schemes, and push them to the corresponding execution entities. The closed-loop verification and iteration module is used to track the execution effect of the optimized decision-making scheme, complete the evaluation closed loop based on the optimized indicator data, and dynamically update the evaluation system and optimization strategy library.

9. The mobile application platform optimization decision-making system based on evaluation results according to claim 8, characterized in that: The comprehensive evaluation module has a built-in weighted comprehensive scoring algorithm unit, which executes the algorithm formula described in claim 2 to calculate the comprehensive evaluation score; the root cause localization module has a built-in coupled root cause analysis algorithm unit, which executes the algorithm formula described in claim 3 to locate the core influencing factors.

Images