Year: 2025
Type: conference-paper
Abstract: Counterfactual evaluation provides a promising framework for assessing explanation fidelity in recommender systems, but perturbation metrics adapted from computer vision suffer three key limitations: (1) they conflate explaining and contradictory features, (2) they average over entire user histories instead of prioritizing concise, high-impact explanations, and (3) they use fixed-percentage perturbations, leading to inconsistencies across users.We introduce refined counterfactual metrics that focus on the most relevant explaining features, exclude contradictory elements, and assess fidelity at a fixed explanation length, ensuring a more consistent and interpretable evaluation.Our code is at: https:// github.com/DeltaLabTLV/FidelityMetrics4XRec
Authors: Mikhail Baklanov, Veronika Bogina, Yehonatan Elisha, Yahlly Schein, Liron Allerhand, Oren Barkan, Noam Koenigstein
Institutions: Tel Aviv University, Open University of Israel
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Funder: Universitas Brawijaya
Award: G1238986628
OpenAlex ID: W4412378278