Deriving user Interest by Mining User Navigation Patterns

摘要

Objectives: The paper focuses on mining the usage patterns to help the user find the pages easily while visiting the website. In order to carry out discovery patterns first the patterns are created and then analyzed for the performance issues. Methods/Statistical Analysis: The quality of a website can be measured by more than one ways, one can consider the aesthetic view of the pages, others can view how the pages are ordered to make it convenient for the user to navigate. We focus on the second perspective of the quality, as it aims at helping the user what it want to find while visiting the website.In order to know what the user wants when he visits the website, it is important to know their browsing patterns. Findings: A lot of good information can be derived from mining log files generated as part of user access to the website. Log files can be mined for basically two reasons 1) To predict the pages that user may prefer to visit and 2) To know the core reason for visiting the site. In order to know the motive of the user visiting the website, it is important to build the browsing patterns of the user. Application/Improvements: In this paper, we present a new way of building patterns to derive user interest by applying predefined classes and assigning weights to the pages accessed during the visit of the user. The algorithm shows better accuracy in predicting the web pages that can be referred by the user in future while navigating the site.