The Design And Implementation Of Online Analysis System Of Crash Information For Android Apps

In recent years,as mobile Internet industry continues to grow rapidly,much atten-tion has been paid to the quality of mobile applications.The frequent update of mobile apps makes it difficult to fully test the apps and in turn causes quality issues.Crash is one of the most serious quality issues in mobile apps.However,crashes on mobile endpoints often fail to provide accurate feedbacks to developers in a timely manner.As a result,developers may miss the best repair time,resulting in the loss of users and economic loss.In order to address the above-mentioned issues in the process of mobile applica-tion crash collection and analysis,this thesis analyzes the real crash instances in mobile applications,and further designs and implements the crash online analysis system for Android mobile applications that integrates the requirements of Android developers.The system consists of three modules:(1)The crash collection module is deployed in the target Android application,for the purpose of monitoring the running state of the application.This module automatically records the application page tracking data and crash screenshots using AOP.In addition,the crash information of application runtime exception is captured by the Android exception handling mechanism(UncaughtEx-ceptionHandler).(2)The crash processing module is deployed on the online serv-er to parse,deduplicate,and classify the crash data collected by the crash collection module.This module implements crash deduplication functionality by employing pat-tern matching and inconsistency analysis method.This module uses naive Bayes and support vector machine,two classification algorithms that are suitable for text clas-sification,to implement the classification of crash data.(3)The crash visualization module is deployed on the online server.This module uses Flask framework to imple-ment back-end functionalities.The front-end uses Antd framework to generate visual crash reports from multiple dimensions,such as stack information,page tracking,de-vice model,system version,and application version.In addition,this module supports real-time data analysis for monitoring application health status.The online crash analysis system fully adapts to the mainstream categories of An-droid applications in the market.Extensive experiments have been performed for 10 different open-source applications 20 Android devices.Exper:imental results demon-strate that,the system implemented in this thesis not only can automatically capture Android application crashes,but also can effectively classify,deduplicate and visualize crash information.The classification accuracy rate achieves 88.1%,and deduplication rate achieves 60.7%.To summarize,this system provides an one-stop crash monitor-ing and analysis solution for the development and testing of Android applications that overcomes the shortcomings of existing crash collection and analysis methods,and improves the quality of mobile applications from all perspectives.