Abstract
In order to address the data sparsity problem in recommender systems, in recent years, Cross-Domain Recommendation (CDR) leverages the relatively richer information from a source domain to improve the recommendation performance on a target domain with sparser information. However, each of the two domains may be relatively richer in certain types of information (e.g., ratings, reviews, user profiles, item details, and tags), and thus, if we can leverage such information well, it is possible to improve the recommendation performance on both domains simultaneously (i.e., dual-target CDR), rather than a single target domain only. To this end, in this paper, we propose a new framework, DTCDR, for Dual-Target Cross-Domain Recommendation. In DTCDR, we first extensively utilize rating and multi-source content information to generate rating and document embeddings of users and items. Then, based on Multi-Task Learning (MTL), we design an adaptable embedding-sharing strategy to combine and share the embeddings of common users across domains, with which DTCDR can improve the recommendation performance on both richer and sparser (i.e., dual-target) domains simultaneously. Extensive experiments conducted on real-world datasets demonstrate that DTCDR can significantly improve the recommendation accuracies on both richer and sparser domains and outperform the state-of-the-art single-domain and cross-domain approaches.
Original language | English |
---|---|
Title of host publication | CIKM 2019 - Proceedings of the 28th ACM International Conference on Information and Knowledge Management |
Subtitle of host publication | Proceedings of the 28th ACM International Conference on Information and Knowledge Management |
Place of Publication | New York |
Publisher | Association for Computing Machinery (ACM) |
Pages | 1533-1542 |
Number of pages | 10 |
ISBN (Electronic) | 9781450369763 |
DOIs | |
Publication status | Published - 2019 |
Event | 28th ACM International Conference on Information and Knowledge Management, CIKM 2019 - Beijing, China Duration: 3 Nov 2019 → 7 Nov 2019 |
Conference
Conference | 28th ACM International Conference on Information and Knowledge Management, CIKM 2019 |
---|---|
Country/Territory | China |
City | Beijing |
Period | 3/11/19 → 7/11/19 |
Keywords
- Recommender Systems
- Cross-Domain Recommendation
- Collaborative Filtering
- Multi-task Learning