Contrastive Learning for Cold-Start Recommendation

  title={Contrastive Learning for Cold-Start Recommendation},
  author={Yin-wei Wei and Xiang Wang and Qi Li and Liqiang Nie and Yan Li and Xuanping Li and Tat-Seng Chua},
  journal={Proceedings of the 29th ACM International Conference on Multimedia},
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.gā€¦Ā 

Figures and Tables from this paper

Socially-aware Dual Contrastive Learning for Cold-Start Recommendation

A socially-aware dual contrastive learning for cold-start recommendation, where cold users can be modeled in the same way as warm users, and a dual-branch self-supervised contrastive objective to account for user-item collaborative features and item-item mutual information.

Task Aligned Meta-learning based Augmented Graph for Cold-Start Recommendation

A fine-grained task aligned task aligned constructor is proposed to cluster similar users and divide tasks for meta-learning, enabling consistent optimization direction, and an augmented graph neural network with two graph enhanced approaches is designed to alleviate data sparsity and capture the high-order user-item interactions.

Improving Item Cold-start Recommendation via Model-agnostic Conditional Variational Autoencoder

This paper attempts to tackle the item cold-start problem by generating enhanced warmed-up ID embeddings for cold items with historical data and limited interaction records using a model-agnostic Conditional Variational Autoencoder based Recommendation(CVAR) framework.

Episodes Discovery Recommendation with Multi-Source Augmentations

This work builds upon the classical Two-Tower model and introduces the novel Multi-Source Augmentations using a Contrastive Learning framework (MSACL) to enhance episodes embedding learning by incorporating positive episodes from numerous correlated semantics.

Generative Adversarial Framework for Cold-Start Item Recommendation

This work proposes a general framework named Generative Adversarial Recommendation (GAR), which can have similar distribution as the warm embeddings that can even fool the recommender, and has strong overall recommendation performance in cold-starting both the CF-based model (improved by over 30.18%) and the GNN- based model ( improved by over 17.78%).

Self-Supervised Learning for Recommender System

This tutorial aims to provide a systemic review of existing self-supervised learning frameworks and analyze the corresponding challenges for various recommendation scenarios, such as general collaborative filtering paradigm, social recommendation, sequential recommendation, and multi-behavior recommendation.

XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation

It is revealed that CL enhances recommendation through endowing the model with the ability to learn more evenly distributed user/item representations, which can implicitly alleviate the pervasive popularity bias and promote long-tail items.

Hierarchical User Intent Graph Network forMultimedia Recommendation

A novel framework, Hierarchical User Intent Graph Network, is developed, which exhibits user intents in a hierarchical graph structure, from the fine-grained to coarse- grained intents, and achieves significant improvements over the state-of-the-art methods, including MMGCN and DisenGCN.

A Survey on Accuracy-oriented Neural Recommendation: From Collaborative Filtering to Information-rich Recommendation

A systematic review on neural recommender models from the perspective of recommendation modeling with the accuracy goal is conducted, aiming to summarize this field to facilitate researchers and practitioners working on recommender systems.

Are Graph Augmentations Necessary?: Simple Graph Contrastive Learning for Recommendation

A simple CL method is proposed which discards the graph augmentations and instead adds uniform noises to the embedding space for creating contrastive views that can smoothly adjust the uniformity of learned representations and has distinct advantages over its graph augmentation-based counterparts in terms of recommendation accuracy and training efficiency.



Meta-learning on Heterogeneous Information Networks for Cold-start Recommendation

This work proposes a novel semantic-enhanced tasks constructor and a co-adaptation meta-learner to address the two questions for how to capture HIN-based semantics in the meta-learning setting, and how to learn the general knowledge that can be easily adapted to multifaceted semantics.

Cold-Start Recommendation with Provable Guarantees: A Decoupled Approach

This paper proposes a novel and general algorithmic framework based on matrix completion that simultaneously exploits the similarity information among users and items to alleviate the cold-start problem and is believed to be the first algorithm that addresses theColdstart problem with provable guarantees on performance.

A Meta-Learning Perspective on Cold-Start Recommendations for Items

This paper proposes two deep neural network architectures that implement a meta-learning strategy to address item cold-start when new items arrive continuously and demonstrates that these techniques significantly beat the MF baseline and also outperform production models for Tweet recommendation.

Item cold-start recommendations: learning local collective embeddings

This work proposes to learn Local Collective Embeddings: a matrix factorization that exploits items' properties and past user preferences while enforcing the manifold structure exhibited by the collective embeddings and presents a learning algorithm based on multiplicative update rules that is efficient and easy to implement.

DropoutNet: Addressing Cold Start in Recommender Systems

This work proposes a neural network based latent model called DropoutNet to address the cold start problem in recommender systems and shows that neural network models can be explicitly trained for cold start through dropout.

Recommendation for New Users and New Items via Randomized Training and Mixture-of-Experts Transformation

A novel model designed to overcome the error superimposition issue and improve model quality is proposed, and extensive experiments show the effectiveness of the proposed model over state-of-the-art alternatives.

Contrastive Pre-training for Sequential Recommendation

A novel model called Contrastive Pre-training for Sequential Recommendation (CP4Rec) is proposed, which utilizes the contrastive pre-training framework to extract meaningful user patterns and further encode the user representation effectively.

Movie genome: alleviating new item cold start in movie recommendation

A new movie recommender system that addresses the new item problem in the movie domain by integrating state-of-the-art audio and visual descriptors and proposing a two-step hybrid approach which trains a CF model on warm items and leverages the learned model on the movie genome to recommend cold items.

MeLU: Meta-Learned User Preference Estimator for Cold-Start Recommendation

A meta-learning-based recommender system called MeLU that can estimate new user's preferences with a few consumed items and provides an evidence candidate selection strategy that determines distinguishing items for customized preference estimation is proposed.

Contrastive Learning for Debiased Candidate Generation in Large-Scale Recommender Systems

CLRec is designed, a contrastive learning method to improve DCG in terms of fairness, effectiveness and efficiency in recommender systems with extremely large candidate size, and improved upon CLRec and proposes Multi-CLRec, for accurate multi-intention aware bias reduction.