Mining association patterns in web usage data
Pang‐Ning Tan · 2002
This paper presents our work in using data mining techniques to discover interesting association patterns from Web usage data. In particular, we address some of the key issues involved in preprocessing and mining the Web association patterns. For preprocessing, we describe how to build accurate classification models to eliminate superfluous sessions created by Web robots. For mining, we develop a new technique called indirect association to capture interesting negative association between different groups of Web users.