TransExplain: Using neural networks to find suitable explanations for Chinese phrases

作者：

Highlights：

• Use the matching method to automatically find the explanation of the phrase.

• Propose a feature extractor covering three levels of word, phrase and sentence.

• Propose a loss function that introduces negative examples into the regression.

• Propose a Chinese phrase and explanation pairs dataset.

摘要

•Use the matching method to automatically find the explanation of the phrase.•Propose a feature extractor covering three levels of word, phrase and sentence.•Propose a loss function that introduces negative examples into the regression.•Propose a Chinese phrase and explanation pairs dataset.

论文关键词：Phrase,Phrase explanation,LSTM,Convolutional neural network,Attention mechanism

论文评审过程：Received 4 May 2020, Revised 11 April 2021, Accepted 12 June 2021, Available online 19 June 2021, Version of Record 24 June 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115440