A memory pool variational autoencoder framework for cross-domain recommendation

Jie Yang, Jianxiang Zhu, Xiaofeng Ding, Yaxin Peng, Yangchun Zhang

Abstract: Highlights•Propose MPVAE to generate embeddings that incorporate more information.•MPVAE involves various relationships among data in the source domain.•MPVAE ensures generation and robustness by the VAE structure.•Conduct extensive experiments on three datasets.