Explainable Recommender Systems with Knowledge Graphs and Language Models

In this tutorial, we delve into recent advances in explainable recommendation using Knowledge Graphs (KGs). The session begins by introducing the fundamental principles behind the increasing adoption of KGs in modern recommender systems. Then, the tutorial explores recent techniques that leverage KGs as an input for language models tailored to explainable recommendation, describing also data types, methods, and evaluation protocols and metrics. Conceptual elements are complemented with hands-on sessions, providing practical implementations using open-source tools and public datasets. Concluding with a comprehensive case study in the education domain as a recap, the tutorial analyses emerging issues and outlines prospective trajectories in this field. The tutorial website is available at https://explainablerecsys.github.io/ecir2024/.