Interpretable Triplet Importance for Personalized Ranking

Personalized item ranking has been a crucial component contributing to the performance of recommender systems. As a representative approach, pairwise ranking directly optimizes the ranking with user implicit feedback by constructing ( user, positive item, negative item ) triplets. Several recent works have noticed that treating all triplets equally may hardly achieve the best effects. They assign different importance scores to negative items, user-item pairs, or triplets, respectively. However, almost all the generated importance scores are groundless and hard to interpret, thus far from trustworthy and transparent. To tackle these, we propose the Triplet Shapley ---a Shapely value-based method to measure the triplet importance in an interpretable manner. Due to the huge number of triplets, we transform the original Shapley value calculation to the Monte Carlo (MC) approximation. To stabilize the MC approximation, we adopt a control covariates-based scheme. Finally, we utilize the triplet Shapley values to guide the resampling of important triplets for facilitating the model learning. Extensive experiments are conducted on six public datasets involving classical matrix factorization- and graph neural network-based recommendation models to demonstrate the superiority of our method.