A new approach to evaluating novel recommendations
Authors:
Open Access
- 23 October 2008
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 179-186
Abstract
This paper presents two methods, named Item- and User-centric, to evaluate the quality of novel recommendations. The former method focuses on analyzing the item-based recommendation network. The aim is to detect whether the network topology has any pathology that hinders novel recommendations. The latter, user-centric evaluation, aims at measuring users' perceived quality of novel, previously unknown, recommendations. The results of the experiments, done in the music recommendation context, show that last.fm social recommender, based on collaborative filtering, is prone to popularity bias. This has direct consequences on the topology of the item-based recommendation network. Pure audio content-based methods (CB) are not affected by popularity. However, a user-centric experiment done with 288 subjects shows that even though a social-based approach recommends less novel items than our CB, users' perceived quality is better than those recommended by a pure CB method.Keywords
This publication has 15 references indexed in Scilit:
- Being accurate is not enough2006
- Experimental Study of Inequality and Unpredictability in an Artificial Cultural MarketScience, 2006
- An industrial-strength content-based music recommendation system2005
- Improving recommendation lists through topic diversification2005
- Interest-based Recommendation in Digital LibraryJournal of Computer Science, 2005
- Evaluating collaborative filtering recommender systemsACM Transactions on Information Systems, 2004
- Assortative Mixing in NetworksPhysical Review Letters, 2002
- Novelty and redundancy detection in adaptive filtering2002
- Item-based collaborative filtering recommendation algorithms2001
- Power-Law Distribution of the World Wide WebScience, 2000