Web Usage Mining has become an important area of research with the rapid expansion of the world wide web. As more data are becoming available, there is much need to study web user behaviour in order to better serve the users and increase business intelligence. In that context, it becomes more important to analyse the next action of the web user. In web usage mining process, web log data is very important for extracting the hidden pattern and discovering the hidden rule in between access for analyzing purpose. It shows some discovery of association rule that might help the website administrator to improve the website structure so that the web pages could be available for user on first attempt. The simulation has been performed on the EPA log file of user behaviour and characteristics of these data sets discovered. In the present work, in web usage mining association rules mining techniques has been used. For this purpose, RapidMiner platform consisting of various web usage mining operators has been used. Analysis of the results showed that using the RapidMiner in web usage mining, can model the frequency of the user visiting the websites. Also, the rules for managing and optimizing the website structure can be achieved.