Greedy Recommending Is Not Always Optimal

Abstract. Recommender systems suggest objects to users. One form recommends documents or other objects to users searching information on a web site. A recommender system can be used to analyse user data to recommend information, for example web pages. Current methods for recommending are aimed at optimising the quality of single recommendations. Even with recommendations, usually many interactions are needed to find the desired information. This changes the task to what we call “interactive recommending”: a series of recommendations in which the user indicates his preference leads to a target object. Here we argue that in interactive recommending a series of normal, “greedy”, recommendings is not the strategy that minimises the number of steps in the search. Greedy sequential recommending conflicts with the need to explore the entire space and may lead to recommending series that require more steps (mouse clicks) from the user than necessary. We illustrate this with an exampl...

• Type: Conference
• Year: 2003
• Where: EWMF
• Authors: Maarten van Someren, Vera Hollink, Stephan ten Hagen
• Keywords: EWMF 2003, Form Recommends Documents, Recommender System, Systems Suggest Objects