User Behavior Analysis In Software Version Management System

At present,the software version management system has a single structural layout,and repeating tedious operations waste a lot of employees' working time,resulting in a poor employee access experience.At the same time,software versions are often delayed,and the company has no objective reference to locate and analyze the causes of the abnormal versions.It is found that the causes of the current situation are closely related to the employees' behavior accessing to the system.Therefore,this paper introduces user behavior analysis technology,digs and analyzes a series of behavior characteristics of employees accessing to the system,grasps employee access needs and working conditions,and applies the analysis results to solve the current problems of the software version management system.This paper proposes the need to analyze employees' behavior by analyzing employee access logs to achieve the goal of improving the status of the software version management system.The paper adopts Hadoop distributed framework,Web log mining technology and cluster analysis algorithms to design and implement a user behavior analysis system for the software version management system.This system is made up for four function modules,log acquisition,log preprocessing,clustering and result visualization application.Among them,the log acquisition module realizes the efficient and unified storage of a large number of employee access logs.The log preprocessing module completes operations such as log cleaning,employee identification,session identification,and path supplementation.The clustering module uses Canopy-based K-means clustering algorithm to perform clustering analysis on logs and group employees based on the analysis results.The results visualization application module applies the grouping results to enhance the employee access experience to the software version management system and help the company locate and analyze the causes of abnormal versions.Finally,the user behavior analysis system is deployed by this paper in a Hadoop cluster environment,where function and performance tests are performed.The test results show that the system can run well,and effectively improve the status of the software version management system,which in line with the design goals and practical application requirements.