A Study of the Similarities of Entity Embeddings Learned from Different Aspects of a Knowledge Base for Item Recommendations.

The recent development of deep learning approaches provides a convenient way to learn entity embeddings from different aspects such as texts and a homogeneous or heterogeneous graph encoded in a knowledge base such as DBpedia. However, it is unclear to what extent domain-specific entity embeddings learned from different aspects of a knowledge base reflect their similarities, and the potential of leveraging those similarities for item recommendations in a specific domain has not been explored. In this work, we investigate domain-specific entity embeddings learned from different aspects of DBpedia with state-of-theart embedding approaches, and the recommendation performance based on the similarities of these embeddings. The experimental results on two real-word datasets show that recommender systems based on the similarities of entity embeddings learned from a homogeneous graph via the dbo: wikiPageWikiLink property provides the best performance compared to the ones learned from other aspects.