Pathwise alignment improves recommendations with sparse feedback
Pathwise Alignment for Collaborative Filtering: Bilateral Conditional Flow Regularization
Information Retrieval
Summary
Recommender systems predict what users might like based on their past choices, but this is hard when they have given very little feedback. The authors propose a new way to train recommendation models called Pathwise Alignment for Collaborative Filtering, which connects users who share some interests by learning paths through their preferences. This approach adds useful constraints from other users’ data without slowing down predictions. They show this method improves recommendation accuracy, especially when user feedback is sparse.
What this means in practice
- •For e-commerce platform teams: Improve product recommendations for users who interact with few items by integrating constraints from other users sharing similar interests.$Commercial implications: Better recommendations can directly improve sales and user engagement, enabling platforms to offer more personalized experiences using sparse customer data.
- •For streaming service engineers: Enhance content suggestions for viewers with limited watch history by aligning user-item preference paths learned from wider user interactions.
Authors
Zhiwei Li, Guodong Long, Chengqi Zhang, Jing Jiang
Abstract
Collaborative filtering predicts user-item interactions from implicit feedback. We focus on matrix factorization (MF) trained with the Bayesian Personalized Ranking loss. This loss updates each user embedding only through that user's observed items and sampled negatives, so sparse feedback gives MF few constraints. To constrain each user through other users, we propose Pathwise Alignment for Collaborative Filtering (PACF), which trains MF with one network shared by all training pairs. At points between user and item embeddings, a flow matching term trains the network to predict the user-to-item step, reversed and scaled for sampled negatives. A trajectory consistency term pulls each user's predicted path to end at the observed item. Both terms update the embeddings, so the network is discarded after training, thereby maintaining the efficiency of MF at inference. We prove that a network with bounded sensitivity to its input point accurately fits users who share an item only if their embeddings are close, so these users constrain one another. Against eleven discriminative and generative baselines, PACF achieves the highest performance on all seven datasets. Its relative gains over MF are largest where feedback is sparsest. The shared network thus supplies the missing constraints through other users.