TY - GEN
T1 - Enhancing Neural Recommender Models through Domain-Specific Concordance
AU - Balashankar, Ananth
AU - Beutel, Alex
AU - Subramanian, Lakshminarayanan
N1 - Publisher Copyright:
© 2021 Owner/Author.
PY - 2021/8/3
Y1 - 2021/8/3
N2 - Recommender models trained on historical observational data alone can be brittle when domain experts subject them to counterfactual evaluation. In many domains, experts can articulate common, high-level mappings or rules between categories of inputs (user's history) and categories of outputs (preferred recommendations). One challenge is to determine how to train recommender models to adhere to these rules. In this work, we introduce the goal of domain-specific concordance: the expectation that a recommender model follow a set of expert-defined categorical rules. We propose a regularization-based approach that optimizes for robustness on rule-based input perturbations. To test the effectiveness of this method, we apply it in a medication recommender model over diagnosis-medicine categories, and in movie and music recommender models, on rules over categories based on movie tags and song genres. We demonstrate that we can increase the category-based robustness distance by up to 126% without degrading accuracy, but rather increasing it by up to 12% compared to baseline models in the popular MIMIC-III, MovieLens-20M and Last.fm Million Song datasets.
AB - Recommender models trained on historical observational data alone can be brittle when domain experts subject them to counterfactual evaluation. In many domains, experts can articulate common, high-level mappings or rules between categories of inputs (user's history) and categories of outputs (preferred recommendations). One challenge is to determine how to train recommender models to adhere to these rules. In this work, we introduce the goal of domain-specific concordance: the expectation that a recommender model follow a set of expert-defined categorical rules. We propose a regularization-based approach that optimizes for robustness on rule-based input perturbations. To test the effectiveness of this method, we apply it in a medication recommender model over diagnosis-medicine categories, and in movie and music recommender models, on rules over categories based on movie tags and song genres. We demonstrate that we can increase the category-based robustness distance by up to 126% without degrading accuracy, but rather increasing it by up to 12% compared to baseline models in the popular MIMIC-III, MovieLens-20M and Last.fm Million Song datasets.
KW - information systems
KW - recommender systems
UR - http://www.scopus.com/inward/record.url?scp=85103004788&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85103004788&partnerID=8YFLogxK
U2 - 10.1145/3437963.3441784
DO - 10.1145/3437963.3441784
M3 - Conference contribution
AN - SCOPUS:85103004788
T3 - WSDM 2021 - Proceedings of the 14th ACM International Conference on Web Search and Data Mining
SP - 1002
EP - 1010
BT - WSDM 2021 - Proceedings of the 14th ACM International Conference on Web Search and Data Mining
PB - Association for Computing Machinery, Inc
T2 - 14th ACM International Conference on Web Search and Data Mining, WSDM 2021
Y2 - 8 March 2021 through 12 March 2021
ER -