Hybrid event recommendation using linked data and user diversity

Khrouf, Houda; Troncy, Raphaël
RECSYS 2013, 7th ACM Recommender Systems Conference, 12-16 October 2013, Hong Kong, China

An ever increasing number of social services o er thousands of diverse events per day. Users tend to be overwhelmed by the massive amount of information available, especially with limited browsing options perceived in many event web services. To alleviate this information overload, a recommender system becomes a vital component for assisting users selecting relevant events. However, such system faces a number of challenges owed to the the inherent complex nature of an event. In this paper, we propose a novel hybrid approach built on top of Semantic Web. On the one hand, we use a content-based system enriched with Linked Data to overcome the data sparsity, a problem induced by the transiency of events. On the other hand, we incorporate a collaborative fi ltering to involve the social aspect, an influential feature in decision making. This hybrid system is enhanced by the integration of a user diversity model designed to detect user propensity towards speci c topics. We show how the hybridization of CB+CF systems and the integration of interest diversity features are important to improve predictions. Experimental results demonstrate the e ectiveness of our approach using precision and recall measures.

Hong Kong
Data Science
Eurecom Ref:
© ACM, 2013. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in RECSYS 2013, 7th ACM Recommender Systems Conference, 12-16 October 2013, Hong Kong, China http://dx.doi.org/10.1145/2507157.2507171

PERMALINK : https://www.eurecom.fr/publication/4127