Discovery of Aggregate Usage Profiles for Web Personalization

TLDR

Web usage mining can enhance personalization by providing actionable patterns, but discovering patterns alone is insufficient; effective derivation of high‑quality, actionable usage is essential. This study introduces and experimentally evaluates two clustering‑based techniques—one on user transactions and one on pageviews—to uncover overlapping aggregate profiles for real‑time recommender systems. The techniques cluster user transactions and pageviews to generate overlapping aggregate profiles, which are then assessed for profile quality and their effectiveness when integrated into a personalization engine.

Abstract

1 Please direct correspondence to [e-mail] Abstract: Web usage mining, possibly used in conjunction with standard approaches to personalization such as collaborative filtering, can help address some of the shortcomings of these techniques, including reliance on subjective user ratings, lack of scalability, and poor performance in the face highdimensional and sparse data. However, the discovery of patterns from usage data by itself is not sufficient for performing the personalization tasks. The critical step is the effective derivation of good quality and useful (i.e., actionable) usage from these patterns. In this paper we present and experimentally evaluate two techniques, based on clustering of user transactions and clustering of pageviews, in order to discover overlapping aggregate profiles that can be effectively used by recommender systems for real-time personalization. We evaluate these techniques both in terms of the quality of the individual profiles generated, as well as in the context of providing recommendations as an integrated part of a personalization engine.