Metrics for Evaluating Explainable Recommender Systems


Hulstijn J., Tchappi I., Najjar A., Aydoğan R.


Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14127 LNAI, pp. 212-230, 2023


Recommender systems aim to support their users by reducing information overload so that they can make better decisions. Recommender systems must be transparent, so users can form mental models about the system’s goals, internal state, and capabilities, that are in line with their actual design. Explanations and transparent behaviour of the system should inspire trust and, ultimately, lead to more persuasive recommendations. Here, explanations convey reasons why a recommendation is given or how the system forms its recommendations. This paper focuses on the question how such claims about effectiveness of explanations can be evaluated. Accordingly, we investigate various models that are used to assess the effects of explanations and recommendations. We discuss objective and subjective measurement and argue that both are needed. We define a set of metrics for measuring the effectiveness of explanations and recommendations. The feasibility of using these metrics is discussed in the context of a specific explainable recommender system in the food and health domain.



Share this page: