Efficient Web Usage Miner Using Decisive Induction Rules

Abstract

Problem statement: Web usage mining is the technique of extracting useful information from server logs (user’s history) and finding out what users are looking for on the Internet. This type of web mining allows for the collection of Web access data for Web pages. Scope: The web usage data provides the paths leading to accessed Web pages with prefrences and higher priorities. This information is often gathered automatically into access logs through the Web server. Approach: In this study we propose Induction based decision rule model for generating inferences and implicit hidden behavioral aspects in the web usage mining which investigates at the web server and client logs. The decision based rule induction mining combines a fast decision rule induction algorithm and a method for converting a decision tree to a simplified rule set. Results: The experimentation is conducted by weka tool and the performance of proposed Induction based decision rule algorithm is evaluated in terms of mined decisive rules, Execution time, root mean square error and mean absolute error. Proposed induction rule mining needs 400 ms of execution time for decisive rule generation, whereas previous work expectation maximization algorithm needs 600ms. Conclusion: Web usage mining is evaluated with decisive rules of user page navigation and preferences. Decisive rule provide the web site developers and owners to known the site presentation likeness and demands of the web users.