A Personalized Recommender System based-on Knowledge Graph Embeddings

“A knowledge graph by means of ontology creates a structured framework for a set of concepts or terms within a specific domain by arranging them in a hierarchical manner, and by using relation descriptors to model the connections between these concepts or terms. This provides a standardized lexicon for representing entities within that domain [11, 16]. Recently, the application of knowledge graphs has grown in popularity within the field of recommender systems and decision support systems for graph-based feature learning [20, 25]”

“By separating features by relation type, we can gain insight into the factors that influence item recommendations for users. As a result, we used these embeddings to evaluate the importance of individual features in relation to the overall results obtained. Additionally, it allows us to identify which relation types are more informative for recommendation and which are less important, and use it to propose an interpretation to users, which can make the recommendation system more transparent, and can also help to improve the user’s understanding and trust of the system.”