Debiased Pairwise Learning for Implicit Collaborative Filtering
- 14 October 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (12) , 7878-7892
Abstract
Learning representations from pairwise comparisons has achieved significant success in various fields, including computer vision and information retrieval. In recommendation systems, collaborative filtering algorithms based on pairwise learning are also rooted in this approach. However, a major challenge in collaborative filtering is the lack of labels for negative instances in implicit feedback data, leading to the inclusion of false negatives among randomly selected instances. This issue causes biased optimization objectives and results in biased parameter estimation. In this paper, we propose a novel method to address learning biases arising from implicit feedback data and introduce a modified loss function for pairwise learning, called debiased pairwise loss (DPL). The core idea of DPL is to correct the biased probability estimates caused by false negatives, thereby adjusting the gradients to more closely approximate those of fully supervised data. Implementing DPL requires only a small modification to the existing codebase. Experimental studies on public datasets demonstrate the effectiveness of the proposed method.Keywords
Funding Information
- National Natural Science Foundation of China (62172167, 52472333)
This publication has 40 references indexed in Scilit:
- Bayesian Negative Sampling for Recommendation2023
- Self-Supervised Learning on Graphs: Contrastive, Generative, or PredictiveIEEE Transactions on Knowledge and Data Engineering, 2021
- SamWalker++: recommendation with informative sampling strategyIEEE Transactions on Knowledge and Data Engineering, 2021
- Understanding the Behaviour of Contrastive Loss2021
- Collaborative List-and-Pairwise Filtering from Implicit FeedbackIEEE Transactions on Knowledge and Data Engineering, 2020
- Momentum Contrast for Unsupervised Visual Representation Learning2020
- Learning from positive and unlabeled data: a surveyMachine Learning, 2020
- Local Aggregation for Unsupervised Learning of Visual Embeddings2019
- Unsupervised Feature Learning via Non-parametric Instance Discrimination2018
- Matrix Factorization Techniques for Recommender SystemsComputer, 2009