Performance Improvement of Web Usage Mining Over Uncertain Data

Abstract: Customer behavior plays an important role in any organization to understand their future need. Web usage mining is an active research topic where customers session clustering is used to understand the customer's activities. It investigates the problem of mining frequent pattern and especially focuses on reducing the number of rules using the partition of a sequential pattern. It also reduces scan size of the database using SOM clustering technique. In the website, every web pages access by the user having some patterns and these patterns are merging and finding the frequent set of web pages. If the user needs next request page in advance then it searches only partial web data, not in whole web data. So it solves the problem through Partition based Pattern Mining Algorithm to remove undetermined data. This system is based on users interest with less execution time. Keywords: Web Usage Mining, Sequence Tree, Web Database, Web Services, Sequential Pattern, Clustering