Website Log Statistical Testing and Analysis

Abstract

In a variety of large data sets, the website log is an important data source. Analyzing the website log, one can get a web page views, viscosity, and related recommendations. The Hadoop ecosystem has the characteristics of open source, active community, covering all aspects of big data processing and a mature ecosystem. In this paper, Hadoop series framework is utilized to test and analyze the website access log. The experiment results demonstrate that the proposed technique can analyze and improve user behavior conveniently.