[论文解读] Self Contrastive Learning for Session-based Recommendation

Zhengxiang Shi, Xi Wang+1 more|arXiv (Cornell University)|Jun 2, 2023

Recommender Systems and Techniques被引用 4

一句话总结

本文提出自对比学习（SCL），一种简化的对比学习方法，通过直接优化项目表征的均匀性，无需构建正负样本或数据增强。SCL在关键指标如P@10和MRR@10上将最先进（SOTA）的会话推荐模型性能提升了8.2%至11.2%，同时通过k-NN近似显著降低了计算复杂度。

ABSTRACT

Session-based recommendation, which aims to predict the next item of users' interest as per an existing sequence interaction of items, has attracted growing applications of Contrastive Learning (CL) with improved user and item representations. However, these contrastive objectives: (1) serve a similar role as the cross-entropy loss while ignoring the item representation space optimisation; and (2) commonly require complicated modelling, including complex positive/negative sample constructions and extra data augmentation. In this work, we introduce Self-Contrastive Learning (SCL), which simplifies the application of CL and enhances the performance of state-of-the-art CL-based recommendation techniques. Specifically, SCL is formulated as an objective function that directly promotes a uniform distribution among item representations and efficiently replaces all the existing contrastive objective components of state-of-the-art models. Unlike previous works, SCL eliminates the need for any positive/negative sample construction or data augmentation, leading to enhanced interpretability of the item representation space and facilitating its extensibility to existing recommender systems. Through experiments on three benchmark datasets, we demonstrate that SCL consistently improves the performance of state-of-the-art models with statistical significance. Notably, our experiments show that SCL improves the performance of two best-performing models by 8.2% and 9.5% in P@10 (Precision) and 9.9% and 11.2% in MRR@10 (Mean Reciprocal Rank) on average across different benchmarks. Additionally, our analysis elucidates the improvement in terms of alignment and uniformity of representations, as well as the effectiveness of SCL with a low computational cost.

研究动机与目标

• 解决现有基于对比学习的会话推荐中对项目表征均匀性利用不足的问题。
• 简化现有依赖复杂正负样本构建与数据增强的复杂对比学习目标。
• 通过一种统一、可解释且可扩展的对比学习目标，仅聚焦于项目表征的均匀性，提升模型性能。
• 评估当SCL应用时，复杂对比学习目标是否仍为必要。
• 通过消融实验与超参数分析，证明SCL的高效性与低计算成本。

提出的方法

• SCL引入一种新颖的对比损失，通过将每个项目的嵌入与所有其他项目进行对比，强制实现所有项目表征的均匀性。
• 核心目标被形式化为自对比损失：$\mathcal{L}^{\text{SCL}} = -\sum_{i=1}^{n} \log \frac{f(\mathbf{x}_i, \mathbf{x}_i)}{\sum_{\mathbf{x}' \in \mathcal{K}_i} f(\mathbf{x}_i, \mathbf{x}')} $，其中$\mathcal{K}_i$为项目$i$的k个最近邻集合。
• 为降低计算成本，该方法采用k-最近邻（kNN）近似，将时间复杂度从$O(n^2 d)$降低至$O(n k d)$。
• SCL被设计为现有SOTA模型中对比目标的即插即用替代方案，无需架构修改或额外的数据增强。
• 该方法兼容任何会话推荐模型，可同时提升表征的对齐性与均匀性。
• 温度超参数$\tau$用于控制对比分布的锐度，最优性能出现在$\tau = 0.1$时。

实验结果

研究问题

• RQ1 SCL在不同基准测试上对会话推荐模型性能的提升程度如何？
• RQ2 SCL在表征对齐性与均匀性方面如何提升模型性能？
• RQ3 当SCL被应用时，复杂的对比学习目标是否仍然必要？
• RQ4 SCL能否在低计算成本下维持最先进性能？

主要发现

• 在不同基准上，SCL使两个SOTA模型在P@10上分别提升8.2%和9.5%，在MRR@10上分别提升9.9%和11.2%。
• SCL在三个数据集（Tmall、Nowplaying和Diginetica）上均一致提升模型性能，且具有统计显著性。
• 使用kNN近似时，负样本数量$k = 2$即达最优性能，表明复杂负采样并非必需。
• 最优温度超参数$\tau$被确定为0.1，更高值会导致性能下降。
• 即使使用完整的负样本集，SCL仍保持强性能，但当$k > 10$后性能趋于平稳，表明收益递减。
• kNN近似显著降低了计算成本，实现了高效部署且不损失性能。