PSL: Rethinking and Improving Softmax Loss from Pairwise Perspective for Recommendation

2024-10-31 · Weiqin Yang, Jiawei Chen, Xin Xin, Sheng Zhou, Binbin Hu, Yan Feng, Chun Chen, Can Wang

Softmax Loss (SL) is widely applied in recommender systems (RS) and has demonstrated effectiveness. This work analyzes SL from a pairwise perspective, revealing two significant limitations: 1) the relationship between SL and conventional ranking metrics like DCG is not sufficiently tight; 2) SL is highly sensitive to false negative instances. Our analysis indicates that these limitations are primarily due to the use of the exponential function. To address these issues, this work extends SL to a new family of loss functions, termed Pairwise Softmax Loss (PSL), which replaces the exponential function in SL with other appropriate activation functions. While the revision is minimal, we highlight three merits of PSL: 1) it serves as a tighter surrogate for DCG with suitable activation functions; 2) it better balances data contributions; and 3) it acts as a specific BPR loss enhanced by Distributionally Robust Optimization (DRO). We further validate the effectiveness and robustness of PSL through empirical experiments. The code is available at https://github.com/Tiny-Snow/IR-Benchmark.

📄 PDF Abstract BibTeX arXiv:2411.00163

Code (1)

tiny-snow/ir-benchmark 공식 구현 pytorch

Tasks

Recommendation Systems

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

GB-CosFace: Rethinking Softmax-based Face Recognition from the Perspective of Open Set Classification

2021-11-22 · Lizhe Liu, Mingqiang Chen, Xiaohao Chen, Siyu Zhu 외

State-of-the-art face recognition methods typically take the multi-classification pipeline and adopt the softmax-based loss for optimization. Although these methods have achieved great success, the softmax-based loss has…

Classification Face Recognition open-set classification TAR

Revisiting Recommendation Loss Functions through Contrastive Learning (Technical Report)

2023-12-13 · Dong Li, Ruoming Jin, Bin Ren

Inspired by the success of contrastive learning, we systematically examine recommendation losses, including listwise (softmax), pairwise (BPR), and pointwise (MSE and CCL) losses. In this endeavor, we introduce InfoNCE+,…

Contrastive Learning

On the Effectiveness of Sampled Softmax Loss for Item Recommendation

2022-01-07 · Jiancan Wu, Xiang Wang, Xingyu Gao, Jiawei Chen 외

The learning objective plays a fundamental role to build a recommender system. Most methods routinely adopt either pointwise or pairwise loss to train the model parameters, while rarely pay attention to softmax loss due …

Contrastive Learning Recommendation Systems Self-Supervised Learning

(Debiased) Contrastive Learning Loss for Recommendation (Technical Report)

2023-12-13 · Ruoming Jin, Dong Li

In this paper, we perform a systemic examination of the recommendation losses, including listwise (softmax), pairwise(BPR), and pointwise (mean-squared error, MSE, and Cosine Contrastive Loss, CCL) losses through the len…

Contrastive Learning

Are LLM-based Recommenders Already the Best? Simple Scaled Cross-entropy Unleashes the Potential of Traditional Sequential Recommenders

2024-08-26 · Cong Xu, Zhangchi Zhu, Mo Yu, Jun Wang 외

Large language models (LLMs) have been garnering increasing attention in the recommendation community. Some studies have observed that LLMs, when fine-tuned by the cross-entropy (CE) loss with a full softmax, could achie…

Sequential Recommendation