Exploiting the semantic similarity of interests in a semantic interest graph for social recommendations
Guangyuan Piao · 2016
Social recommendation, a recommender system that targets social media domains, has attracted increasing attention with the growing popularity of Online Social Networks (OSNs). User Interest Modeling (UIM) and Recommendation Algorithm (RA) as two major components play significant roles in such a system. In recent studies, the Semantic Interest Graph (SIG), which represents the interests of a user as resources in DBpedia, has shown the efficiency on UIM in OSNs. However, the similarity between resources is not exploited when using SIG in RAs. In this work, we propose a novel semantic similarity measure for calculating the similarity between resources. We show preliminary results on the performance of calculating the similarity between general resources as well as resources in a single domain for social recommendations.