Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach

Qitian Wu, Hengrui Zhang, Xiaofeng Gao, Junchi Yan, Hongyuan Zha
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:11329-11339, 2021.

Abstract

Recommendation models can effectively estimate underlying user interests and predict one’s future behaviors by factorizing an observed user-item rating matrix into products of two sets of latent factors. However, the user-specific embedding factors can only be learned in a transductive way, making it difficult to handle new users on-the-fly. In this paper, we propose an inductive collaborative filtering framework that contains two representation models. The first model follows conventional matrix factorization which factorizes a group of key users’ rating matrix to obtain meta latents. The second model resorts to attention-based structure learning that estimates hidden relations from query to key users and learns to leverage meta latents to inductively compute embeddings for query users via neural message passing. Our model enables inductive representation learning for users and meanwhile guarantees equivalent representation capacity as matrix factorization. Experiments demonstrate that our model achieves promising results for recommendation on few-shot users with limited training ratings and new unseen users which are commonly encountered in open-world recommender systems.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-wu21j, title = {Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach}, author = {Wu, Qitian and Zhang, Hengrui and Gao, Xiaofeng and Yan, Junchi and Zha, Hongyuan}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {11329--11339}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/wu21j/wu21j.pdf}, url = {https://proceedings.mlr.press/v139/wu21j.html}, abstract = {Recommendation models can effectively estimate underlying user interests and predict one’s future behaviors by factorizing an observed user-item rating matrix into products of two sets of latent factors. However, the user-specific embedding factors can only be learned in a transductive way, making it difficult to handle new users on-the-fly. In this paper, we propose an inductive collaborative filtering framework that contains two representation models. The first model follows conventional matrix factorization which factorizes a group of key users’ rating matrix to obtain meta latents. The second model resorts to attention-based structure learning that estimates hidden relations from query to key users and learns to leverage meta latents to inductively compute embeddings for query users via neural message passing. Our model enables inductive representation learning for users and meanwhile guarantees equivalent representation capacity as matrix factorization. Experiments demonstrate that our model achieves promising results for recommendation on few-shot users with limited training ratings and new unseen users which are commonly encountered in open-world recommender systems.} }
Endnote
%0 Conference Paper %T Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach %A Qitian Wu %A Hengrui Zhang %A Xiaofeng Gao %A Junchi Yan %A Hongyuan Zha %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-wu21j %I PMLR %P 11329--11339 %U https://proceedings.mlr.press/v139/wu21j.html %V 139 %X Recommendation models can effectively estimate underlying user interests and predict one’s future behaviors by factorizing an observed user-item rating matrix into products of two sets of latent factors. However, the user-specific embedding factors can only be learned in a transductive way, making it difficult to handle new users on-the-fly. In this paper, we propose an inductive collaborative filtering framework that contains two representation models. The first model follows conventional matrix factorization which factorizes a group of key users’ rating matrix to obtain meta latents. The second model resorts to attention-based structure learning that estimates hidden relations from query to key users and learns to leverage meta latents to inductively compute embeddings for query users via neural message passing. Our model enables inductive representation learning for users and meanwhile guarantees equivalent representation capacity as matrix factorization. Experiments demonstrate that our model achieves promising results for recommendation on few-shot users with limited training ratings and new unseen users which are commonly encountered in open-world recommender systems.
APA
Wu, Q., Zhang, H., Gao, X., Yan, J. & Zha, H.. (2021). Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:11329-11339 Available from https://proceedings.mlr.press/v139/wu21j.html.

Related Material