An Autoencoder Framework With Attention Mechanism for Cross-Domain Recommendation
Philip S. Yu,J. Lai,Changdong Wang,Ling Huang,Shi-Ting Zhong
DOI: https://doi.org/10.1109/TCYB.2020.3029002
IF: 11.8
2020-11-06
IEEE Transactions on Cybernetics
Abstract:In recent years, the recommender system has been widely used in online platforms, which can extract useful information from giant volumes of data and recommend suitable items to the user according to user preferences. However, the recommender system usually suffers from sparsity and cold-start problems. Cross-domain recommendation, as a particular example of transfer learning, has been used to solve the aforementioned problems. However, many existing cross-domain recommendation approaches are based on matrix factorization, which can only learn the shallow and linear characteristics of users and items. Therefore, in this article, we propose a novel autoencoder framework with an attention mechanism (AAM) for cross-domain recommendation, which can transfer and fuse information between different domains and make a more accurate rating prediction. The main idea of the proposed framework lies in utilizing autoencoder, multilayer perceptron, and self-attention to extract user and item features, learn the user and item-latent factors, and fuse the user-latent factors from different domains, respectively. In addition, to learn the affinity of the user-latent factors between different domains in a multiaspect level, we also strengthen the self-attention mechanism by using multihead self-attention and propose AAM++. Experiments conducted on two real-world datasets empirically demonstrate that our proposed methods outperform the state-of-the-art methods in cross-domain recommendation and AAM++ performs better than AAM on sparse and large-scale datasets.
Medicine,Computer Science