The Design And Implementation Of User Behavior Visualization Analysis For Access Path

With the growing popularity of the website, more and more people are startingto focus on website traffic and customer groups and their need s,user access pathanalys is which is an important aspect of traffic analysis has also attracted more andmore users’ attention. The administrators of websites need to know exactly howmany people visit their site, where did they come from,what part and services of thesite are the user interested in, what are their needs,and so on. Therefore, large siteshave their own flow statistics analys is system. In order to accurately understand theweb sites vis ited, customer groups, the website and any page’s dynamic flow and it’spopularity, and thus make the relevant adjustment push strategy to achieve bettercustomer service.Traffic statistics is one of the important way to improve of Web services, anduser behavior analysis is the important part of Web site traffic statistics, users usepath analysis access can intuitively reflect the user habit. Through the analys is of theuser access path, we can understand the relationship between how each page jump,what areusers more accustomed to access the site. From all above, we can tagert tooptimize the web site.This paper focuses on the user access path. Server-side uses Hadooparchitecture; the upper use hive to combine the Hadoop and data warehouse and weuse ETL to process data before data storage. After obtaining the user’s access logs,first we create an index of the data dump, and then save the user access paths foreach product line and page for each category, fina lly build user access path treesbetween every two nodes. When users use this tool, they need first select a filteroption and feed the selecting back to the server-side, server-side query the databasewith the SQL which generate by the users selecting, and then server find therevelvant user access tree to the front and then use canvas to draw the tree. In theprocess, we use Apriori algorithm to calculate the path of the tree and displayeddirectly in the path of the tree, and thus to optimize web pages and websites.