Web Navigation Path Pattern Prediction Using First Order Markov Model and Depth First Evaluation
Web usage mining has been defined as a technique of finding hidden knowledge from a log file. The interaction between website and user is recorded in the related web server log file. Web designer is able to analyze the file in order to understand the interaction between users and a web site, which helps to improve web topology. All information of web usage can be generated from log files and it consists of set of navigation sessions that represent the trails formed by users during the navigation process. In this paper, user web navigation sessions are inferred from log data and are modeled as a Markov chain. The chain's higher probability trails will be the most likely preferred trails on the web site.