Heterogeneous deep graph convolutional network with citation relational BERT for COVID-19 inline citation recommendation

作者：

Highlights：

• A neural based citation recommendation method is proposed for COVID-19 papers.

• A new solution to retrain and fine-tune PLMs is proposed.

• A new heterogeneous deep graph convolutional network is proposed.

• CRB-HDGCN achieves state-of-art on CORD-19 and LitCovid datasets.

摘要

•A neural based citation recommendation method is proposed for COVID-19 papers.•A new solution to retrain and fine-tune PLMs is proposed.•A new heterogeneous deep graph convolutional network is proposed.•CRB-HDGCN achieves state-of-art on CORD-19 and LitCovid datasets.

论文关键词：COVID-19 citation recommendation,Deep graph convolutional network,Heterogeneous graph,Citation enhanced BERT,Text representation learning

论文评审过程：Received 12 July 2022, Revised 2 September 2022, Accepted 12 September 2022, Available online 17 September 2022, Version of Record 27 September 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118841