Exploiting Semantic Distance in Linked Open Data for Recommendation

The use of Linked Open Data (LOD) has been explored in recommender systems in different ways, primarily through its graphical representation. The graph structure of LOD is utilized to measure inter-resource relatedness via their semantic distance in the graph. The intuition behind this approach is that the more connected resources are to each other, the more related they are. One drawback of this approach is that it treats all inter-resource connections identically rather than prioritizing links that may be more important in semantic relatedness calculations. Another drawback of current approaches is that they only consider resources that are connected directly or indirectly through an intermediate resource only. In this document, we show that different types of inter-resource links hold different values for relatedness calculations between resources, and we exploit this observation to introduce improved resource semantic relatedness measures that are more accurate than the current state of the art approaches. Moreover, we introduce an approach to propagate current semantic distance approaches that does not only expand the coverage of current approaches, it also increases their accuracy. To validate the effectiveness of our approaches, we conducted several experiments to identify the relatedness between musical artists in DBpedia, and they demonstrated that approaches that prioritize link types resulted in more accurate recommendation results. Also, propagating semantic distances beyond one hub resources does not only result in an improved accuracy, it also shows that propagating semantic distances beyond one hub resources improves the coverage of LOD-based recommender