Improving Micro-video Recommendation via Contrastive Multiple Interests

@article{Li2022ImprovingMR,
  title={Improving Micro-video Recommendation via Contrastive Multiple Interests},
  author={Beibei Li and Beihong Jin and Jiageng Song and Yisong Yu and Y. Zheng and Wei Zhuo},
  journal={Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval},
  year={2022}
}
  • Beibei Li, Beihong Jin, Wei Zhuo
  • Published 19 May 2022
  • Computer Science
  • Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval
With the rapid increase of micro-video creators and viewers, how to make personalized recommendations from a large number of candidates to viewers begins to attract more and more attention. However, existing micro-video recommendation models rely on expensive multi-modal information and learn an overall interest embedding that cannot reflect the user's multiple interests in micro-videos. Recently, contrastive learning provides a new opportunity for refining the existing recommendation… 

Figures and Tables from this paper

References

SHOWING 1-10 OF 22 REFERENCES
User-Video Co-Attention Network for Personalized Micro-video Recommendation
TLDR
A novel framework User-Video Co-Attention Network (UVCAN) is proposed, which can learn multi-modal information from both user and microvideo side using attention mechanism and reasons about the attention in a stacked attention network fashion for both users and micro-video.
MMGCN: Multi-modal Graph Convolution Network for Personalized Recommendation of Micro-video
TLDR
A Multi-modal Graph Convolution Network (MMGCN) framework built upon the message-passing idea of graph neural networks, which can yield modal-specific representations of users and micro-videos to better capture user preferences is designed.
Temporal Hierarchical Attention at Category- and Item-Level for Micro-Video Click-Through Prediction
TLDR
This paper investigates how to model user's historical behaviors so as to predict the user's click-through of micro-videos, and proposes a Temporal Hierarchical Attention at Category- and Item-Level (THACIL) network for user behavior modeling.
What Aspect Do You Like: Multi-scale Time-aware User Interest Modeling for Micro-video Recommendation
TLDR
This work proposes an end-to-end Multi-scale Time-aware user Interest modeling Network (MTIN), which presents an interest group routing algorithm to generate fine-grained user interest groups based on user's interaction sequence and achieves substantial gains over the state-of-the-art methods.
Controllable Multi-Interest Framework for Recommendation
TLDR
This paper proposes a novel controllable multi-interest framework for the sequential recommendation, called ComiRec, which captures multiple interests from user behavior sequences, which can be exploited for retrieving candidate items from the large-scale item pool.
Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation
TLDR
Visualization results of the learned representations validate that DuoRec can largely alleviate the representation degeneration problem, and a model-level augmentation is proposed based on Dropout to enable better semantic preserving.
Graph Augmentation-Free Contrastive Learning for Recommendation
TLDR
A graph augmentation-free CL method to simply adjust the uniformity of the learned user/item representation distributions on the unit hypersphere by adding uniform noises to the original representations for data augmentations, and enhance recommendation from a geometric view is proposed.
Octopus: Comprehensive and Elastic User Representation for the Generation of Recommendation Candidates
TLDR
A novel personalized candidate generation paradigm, Octopus, is proposed, which is remarkable for its comprehensiveness and elasticity, and will not only identify user's interested items comprehensively, but also rule out irrelevant candidates and help to maintain a feasible running cost.
Contrastive Self-supervised Sequential Recommendation with Robust Augmentation
TLDR
This paper introduces two informative augmentation operators leveraging item correlations to create high quality views for contrastive learning, and proposes a novel framework, Contrastive Self-supervised Learning for Sequential Recommendation (CoSeRec), on improving model performance, and the robustness against sparse and noisy data.
The World is Binary: Contrastive Learning for Denoising Next Basket Recommendation
TLDR
This paper proposes a Contrastive Learning Model~(named CLEA) to automatically extract items relevant to the target item for next basket recommendation and demonstrates that the proposed CLEA achieves significantly better recommendation performance than the existing state-of-the-art alternatives.
...
...