Recommendation Agents on the Internet

Abstract

This article examines one of the functions of "intelligent agents" used by service providers on the Internet (e.g., retailers and information search engines) that compare a user's profile to data on other users to determine which users in the database are similar to the focal user, and which recommendations (of products or information strings) might thereby be useful to the focal user. We characterize these questions as Internet modelers "rediscovering the wheel" of cluster analysis. We draw from the literature on cluster analysis to begin to address the questions being posed in this new application arena.