Semantic context-aware recommendation via topic models leveraging linked open data

Abstract

Context aware recommendation systems are used to provide personalized recommendations by exploiting contextual situation. They take into account not only user preferences,but also additional relevant information (context). Statistical topic models such as Latent Dirichlet Allocation (LDA) have been extensively used for discovering latent semantic topics in text documents. In this paper,we propose a probabilistic topic model that incorporates user interests,item representation and context information in a single framework. In our approach,the contextual information is represented as a subset of the items feature space which is acquired from the knowledge available in the Linked Open Data (LOD). We use DBpedia,a well-known knowledge base in LOD,to utilize the context information in recommendation. Our proposed recommendation framework computes the conditional probability of each item given the user preferences and the additional context. We use these probabilities as recommendation scores to find top-n items for recommendations. The performed experiments demonstrate the effectiveness of our proposed method and shows that leveraging semantic context from the Linked Open Data can improve the quality of the recommendations.