Discovering Generalized Association Rule in Web Usage Mining by Frequent Pattern Tree (FP-TREE)

Abstract

Web mining techniques can be used to search for web access patterns, web structures, regularity and dynamics of web contents. Web usage mining analyzes Web log files to discover user accessing patterns of Web pages. Log file data can offer valuable insight into web site usage. It reflects actual usage in natural working condition, compared to the artificial setting of a usability lab. This system presents web log mining based on hierarchy of web usage data by generalized association rule. Multi-level association rule is used for implementation of generalized association rule. In this system, Web log database is used to store web access records in log files collected from web server. And web log databases are constructed via a process of data cleaning, data transformation. By using Frequent Pattern Tree (FP- Tree), the system generates rules from web log data, and reduces counting phase of association rule since it stores the pre-computed count values. Frequent patterns are generated instead of page item-sets. The generated frequent patterns can later be applied to improve web site management, decision making process. This system is implemented using ASP.Net programming language with Microsoft SQL Server 2008 R2 Database Engine.