Legal case document similarity: You need both network and text

作者：

Highlights：

• Estimating the similarity between legal case documents is a practically useful task.

• We propose Hier-SPCNet that substantially improve the network-based similarity.

• We apply several methods for combining textual and network information for the task.

• We demonstrate the utility of our method by recommending similar case documents.

摘要

•Estimating the similarity between legal case documents is a practically useful task.•We propose Hier-SPCNet that substantially improve the network-based similarity.•We apply several methods for combining textual and network information for the task.•We demonstrate the utility of our method by recommending similar case documents.

论文关键词：Legal IR,Legal document similarity,Citation network,Heterogeneous network,Network embeddings,Text embeddings,Combining text and network similarity

论文评审过程：Received 15 March 2022, Revised 8 August 2022, Accepted 19 August 2022, Available online 5 September 2022, Version of Record 5 September 2022.

论文官网地址：https://doi.org/10.1016/j.ipm.2022.103069