Contrastive Learning for Sequential Recommendation

By Xu Xie, Fei Sun, Zhaoyang Liu, Shiwen Wu, Jinyang Gao, Jiandong Zhang, Bolin Ding, Bin Cui. IEEE International Conference on Data Engineering, 2022

559 citations recorded by the source corpus.

Sequential recommendation methods play a crucial role in modern recommender systems because of their ability to capture a user's dynamic interest from her/his historical inter-actions. Despite their success, we argue that these approaches usually rely on the sequential prediction task to optimize the huge amounts of parameters. They usually suffer from the data sparsity problem, which makes it difficult for them to learn high-quality user representations. To tackle that, inspired by recent advances of contrastive learning techniques in the computer vision, we propose a novel multi-task framework called Contrastive Learning for Sequential Recommendation (CL4SRec). CL4SRec not only takes advantage of the traditional next item prediction task but also utilizes the contrastive learning framework to derive self-supervision signals from the original user behavior sequences. Therefore, it can extract more meaningful user patterns and further encode the user representations effectively. In addition, we propose three data augmentation approaches to construct self-supervision signals. Extensive experiments on four public datasets demonstrate that CL4SRec achieves state-of-the-art performance over existing baselines by inferring better user representations.

Subfields: Artificial Intelligence, Information Systems, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/a8f2764a76772ab957b31f5d7bb88527d46fbdc8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Wayne Xin Zhao
2. Hongzhi Yin
3. Ji-rong Wen
4. Xiangnan He
5. Yupeng Hou
6. Tong Chen
7. Chao Huang
8. Xin Xia

In Information Systems

1. Hongzhi Yin
2. Wayne Xin Zhao
3. Tong Chen
4. Junliang Yu
5. Xin Xia
6. Yupeng Hou
7. Xiangnan He
8. Chao Huang

In Management Science and Operations Research

1. Xiangnan He
2. Fuli Feng
3. Wayne Xin Zhao
4. Ji-rong Wen
5. Wenjie Wang
6. Ruihong Qiu
7. Zi-Liang Huang
8. Bolin Ding

Citations in the corpus

• Are Graph Augmentations Necessary?: Simple Graph Contrastive Learning for Recommendation (2021, background): “CL is often applied to recommendation with a particular set of presumed representational invariances to data augmentations [28, 33, 40, 45].”
• TALLRec: An Effective and Efficient Tuning Framework to Align Large Language Model with Recommendation (2023, background): “Thus, some works focus on the generalization ability of sequential recommendation models by pretraining [32, 60], data augmentation [37, 38, 50, 56],”
• Towards Universal Sequence Representation Learning for Recommender Systems (2022, methodology): “Besides directly mining sequential patterns of IDs, some works are proposed to model sequences with rich features [32, 36] or additional self-supervised signals [29, 30, 36].”
• LightGCL: Simple Yet Effective Graph Contrastive Learning for Recommendation (2023, background): “The main idea of contrastive learning in enhancing the user and item representation is to research the agreement between the generated embedding views by contrasting the defined positive pairs with negative instance counterparts (Xie et al., 2022).”
• Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation (2021, methodology): “, item cropping, masking, and reordering [49, 50], which may cause semantically inconsistent samples.”
• Self-Supervised Learning for Recommender Systems: A Survey (2022, methodology): “Some works [58], [59] propose to shuffle a continuous sub-sequence [ i r , i r +1 , . . . , i r + L r − 1 ] to [˜ i r , ˜ i r +1 , . . . , ˜ i r + L r − 1 ] to create sequence augmentations, which is formulated as: (10) Item Substitution: Random item cropping and masking may exaggerate the data…”
• XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation (2022, background): “Generally, data augmentations are the prerequisite for CL-based recommendation models [12], [13], [15], [25].”
• Intent Contrastive Learning for Sequential Recommendation (2022, methodology): “Recent methods [45, 51] in SR follow standard contrastive SSL to maximize mutual information between two positive views of sequences.”
• Don't Contrast the Impossible: Region-Constrained Batching for Contrastive User Modeling on a Local Community Platform (2026, unknown): “Contrastive learning (CL) is now the de facto standard for learning representations from behavioral logs in recommendation systems [22, 23, 26].”