Pattern Discovery of Usage Pattern from Web Data

Web usage mining is a data mining methods. There are huge amount of data are stored on the internet. When user find any particular information by search engine like Google, Bing etc. is very hard because the complexity of web pages is increases day by day. Web usage mining play a vital role to solve this problem. In web usage mining we are creating an appropriate pattern according to the user’s visiting behavior. The aim of this dissertation report is to applying an association rule using FP-Growth algorithm on web server log file (an educational institution web log data) to find the behavioral pattern and profiles of users interacting with a web site. The web mining usage pattern of a Technical Institution web data. Web related data is coteries in to three parts namely web log, access log, and proxy log data and collect the data in web server and we analyze this data using web log analyzer tool and WEKA tool. Our experimental results help to predict and identify the number of visitor for the website and improve the website usability. This work using data from user visiting system web log files that produced from web server IIS (Internet Information-Services) and using the accessed web address page references and access time.