Product recommender systems aim to support consumers in making buying decisions. However, such a support requires considering the consumer behaviour in making buying decisions. In this paper, we deduce design requirements for utility-based recommender systems from the theory of consumer information behaviour and present empirically findings from experiments conducted with a prototypical implementation of the proposed requirements. The empirical examination shows that our recommender system has a high predictive validity.
Scholz, Michael, "Implications of Consumer Information Behaviour to Construct Utility-based Recommender Systems: A Prototypical Study" (2010). BLED 2010 Proceedings. 19.