Contrastive Learning for Cold-Start Recommendation

By Yin-wei Wei, Xiang Wang, Qi Li, Liqiang Nie, Yan Li, Xuanping Li, Tat-seng Chua. ACM Multimedia, 2021

365 citations recorded by the source corpus.

Recommending purely cold-start items is a long-standing and fundamental challenge in the recommender systems. Without any historical interaction on cold-start items, the collaborative filtering (CF) scheme fails to leverage collaborative signals to infer user preference on these items. To solve this problem, extensive studies have been conducted to incorporate side information of items (e.g. content features) into the CF scheme. Specifically, they employ modern neural network techniques (e.g., dropout, consistency constraint) to discover and exploit the coalition effect of content features and collaborative representations. However, we argue that these works less explore the mutual dependencies between content features and collaborative representations and lack sufficient theoretical supports, thus resulting in unsatisfactory performance on cold-start recommendation. In this work, we reformulate the cold-start item representation learning from an information-theoretic standpoint. It aims to maximize the mutual dependencies between item content and collaborative signals. Specifically, the representation learning is theoretically lower-bounded by the integration of two terms: mutual information between collaborative embeddings of users and items, and mutual information between collaborative embeddings and feature representations of items. To model such a learning process, we devise a new objective function founded upon contrastive learning and develop a simple yet efficient Contrastive Learning-based Cold-start Recommendation framework (CLCRec). In particular, CLCRec consists of three components: contrastive pair organization, contrastive embedding, and contrastive optimization modules. It allows us to preserve collaborative signals in the content representations for both warm and cold-start items. Through extensive experiments on four publicly accessible datasets, we observe that CLCRec achieves significant improvements over state-of-the-art approaches in both warm- and cold-start scenarios.

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

View Paper: https://www.semanticscholar.org/paper/2ecbb88131453b9a61bd61320c19d5581106c442

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. Xiangnan He
2. Hongzhi Yin
3. Xiang Wang
4. Chao Huang
5. Junliang Yu
6. Tong Chen
7. Xin Xia
8. Meng Wang

In Information Systems

1. Hongzhi Yin
2. Xiangnan He
3. Xiang Wang
4. Junliang Yu
5. Tong Chen
6. Chao Huang
7. Xin Xia
8. Meng Wang

In Management Science and Operations Research

1. Xiang Wang
2. Meng Wang
3. Xiangnan He
4. Le Wu
5. Kun Zhang
6. Zi-Liang Huang
7. Jundong Li
8. Chao Huang

Citations in the corpus

• LLMRec: Large Language Models with Graph Augmentation for Recommendation (2023, methodology): “Data sparsity and the cold-start problem hinder collaborative preference capturing [39].”
• A Survey on Accuracy-Oriented Neural Recommendation: From Collaborative Filtering to Information-Rich Recommendation (2021, background): “Besides, self-supervised learning [263], [264] is becoming emerged and showing promises in recommendation tasks [265], [266], [267], [268], [269].”
• Are Graph Augmentations Necessary?: Simple Graph Contrastive Learning for Recommendation (2021, background): “[27] reformulated the cold-start item representation learning from an information-theoretic standpoint and maximized the mutual dependencies between item content and collaborative signals to alleviate the data sparsity issue.”
• Self-Supervised Learning for Recommender Systems: A Survey (2022, background): “However, CL can be applied to both supervised and unsupervised settings, and those CL-based recommendation methods which do not augment the raw data [47], [48] and just optimize a marginal loss [49], [50], should not be roughly classified into SSR either.”
• XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation (2022, positive): “Inspired by the success of CL in other fields, the community also has started to integrate CL into recommendation [15], [12], [14], [17], [13], [51], [52], [16].”