How to Surprisingly Consider Recommendations? A Knowledge-Graph-Based Approach Relying on Complex Network Metrics

Abstract

Abstract

en Traditional recommendation proposals, including content-based and collaborative filtering, usually focus on similarity between items or users. Existing approaches lack ways of introducing unexpectedness into recommendations, prioritizing globally popular items over exposing users to unforeseen items. This investigation aims to design and evaluate a novel layer on top of recommender systems suited to incorporate relational information and rerank items with a user-defined degree of surprise. Surprise in recommender systems refers to the degree to which a recommendation deviates from the user’s expectations, providing an unexpected yet relatable recommendation. We propose a knowledge graph-based recommender system by encoding user interactions on item catalogs. Our study explores whether network-level metrics on knowledge graphs (KGs) can influence the degree of surprise in recommendations. We hypothesize that surprisingness correlates with specific network metrics, treating user profiles as subgraphs within a larger catalog KG. The achieved solution reranks recommendations based on their impact on structural graph metrics. Our research contributes to optimizing recommendations to reflect the network-based metrics. We experimentally evaluate our approach on two datasets of LastFM listening histories and synthetic Netflix viewing profiles. We find that reranking items based on complex network metrics leads to a more unexpected and surprising composition of recommendation lists.