Research Of Crash Monitoring And User Behavior Analysis On Mibole Application

Recently years,mobile application develops explosively under the impetus of mobile Internet technology.Mobile application has brought various negative questions while bringing the great convenience to people's lives.The uneven quality of applications,obvious homogenization phenomenon and fierce competition among software developers have result in higher requirements for the developing of mobile application.With the purpose to improve the quality of mobile application as well as enhance the user experience,this paper studies the key techniques about crash monitoring and user behavior analysis of mobile application.This dissertation designs a crash monitoring and user behavior analysis platform by analyzing the pros and cons of recent mobile application testing techniques and clickstream analysis techniques.However,it needs some key technologies to implement the platform,including:(1)design a feasible crash analysis scheme to monitor crashes occur in the process of using Android application,which helps developers recognize crashes timely and fix them quickly.(2)Propose an algorithm for maximal frequent sequence mining,which can discover the user access patterns from large volume of clickstream data,then help developers adjust the organizational structure among pages.(3)Provide a critical path conversion rate analysis algorithm for operators to analyze and improve the conversion rate of critical process.(4)Present a two-level clustering algorithm to devide users into defferent clusters according to access habits of users,which helps operators provide subsequent personalized services.At last,this paper uses a typical application as experimental object to verify the efficiency of crash analysis scheme and user behavior analysis algorithms.