Mining web content usage patterns of electronic commerce transactions for enhanced customer services

Abstract A successful business intelligence solution can help organizations improve the quality and speed of their decision‐making processes by analyzing the consolidated information collected from their websites. Using the current Web server log standard, which indicates only the locations of served Web pages, may lead to inaccurate business analysis for data driven and frequently updated static content Web pages. A properly defined Web content usage data warehouse that captures both dynamic and static contents of web pages provides rich data source for discovering interesting business rules among users' activities. This paper demonstrates the simplicity of data extract, transform and load procedures to import raw Web content usage log to various data models for data analysis, reporting and data mining tools. In this paper, we use two data mining techniques (expected maximization and the prefixspan algorithms) for visitor grouping and path analysis to find interesting patterns in Web content usage log data, an important component of e‐commerce Web sites traffic analysis. Visitor grouping uses data clustering to group visitors/sessions with similar selected attributes' value. Path analysis mines common visiting path sequence. The ultimate goal is to enable the online merchant provide enhanced and differentiated marketing services generally to its existing and potential customers and tell what items customers are interested in instead of ambiguous URLs