DSER: Deep-Sequential Embedding for single domain Recommendation

作者：

Highlights：

• It is the first study using Doc2vec to model sequential interaction and fuse with DNN.

• DSER outperforms the other SOTA approaches in HR and NDCG on four real-world datasets.

• It shows significant superiority on the tourism datasets having a high new user ratio.

• We empirically find out how Doc2vec works with DSER in terms of data characteristics.

摘要

•It is the first study using Doc2vec to model sequential interaction and fuse with DNN.•DSER outperforms the other SOTA approaches in HR and NDCG on four real-world datasets.•It shows significant superiority on the tourism datasets having a high new user ratio.•We empirically find out how Doc2vec works with DSER in terms of data characteristics.

论文关键词：Recommender system,Deep learning,Word embedding,Single-domain recommendation,Smart tourism

论文评审过程：Received 9 February 2022, Revised 30 June 2022, Accepted 11 July 2022, Available online 14 July 2022, Version of Record 21 July 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118156