Unsupervised neural networks for automatic Arabic text summarization using document clustering and topic modeling

作者：

Highlights：

• Document representation in the topic space improves the summarization compared to BOW.

• Feature learning using unsupervised neural networks improves the summarization task.

• Unsupervised neural networks trained on sentence/topic vectors give promising results.

• Ensemble learning with topic representation obtains the best results.

摘要

•Document representation in the topic space improves the summarization compared to BOW.•Feature learning using unsupervised neural networks improves the summarization task.•Unsupervised neural networks trained on sentence/topic vectors give promising results.•Ensemble learning with topic representation obtains the best results.

论文关键词：Arabic text summarization,Natural language processing,Deep learning,Neural networks,Clustering,Topic modeling

论文评审过程：Received 14 May 2020, Revised 16 September 2020, Accepted 21 January 2021, Available online 2 February 2021, Version of Record 13 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114652