Collaborative filtering for producing recommendations in the retail sector

Abstract

Recommender Systems exploit implicit or explicit user feedback, to create recommendations and provide a personalized user experience. In the case of explicit feedback datasets, the system directly collects the user opinion. On the other hand, to compile implicit feedback datasets the system works passively in the background, tracking different sorts of user behavior, such as browsing activity, watching habits or purchase history. In this work, we focus on implicit feedback recommendation systems. We analyze their unique characteristics and identify their differences to the much more extensively researched explicit feedback systems.