An unsupervised method for extractive multi-document summarization based on centroid approach and sentence embeddings
作者：
Highlights：
• An unsupervised method for extractive multi-document summarization.
• Pre-trained sentence embedding models are used for sentences representations.
• Centroid approach is applied to compute the sentence content relevance score.
• Sentence selection based on sentence relevance, novelty and position scores.
• The use of sentence embedding methods leads to significant improvements.
摘要
•An unsupervised method for extractive multi-document summarization.•Pre-trained sentence embedding models are used for sentences representations.•Centroid approach is applied to compute the sentence content relevance score.•Sentence selection based on sentence relevance, novelty and position scores.•The use of sentence embedding methods leads to significant improvements.
论文关键词：Extractive text summarization,Word embeddings,Sentence embeddings,Centroid approach,Transfer learning
论文评审过程：Received 25 October 2019, Revised 3 September 2020, Accepted 22 October 2020, Available online 27 October 2020, Version of Record 10 February 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.114152
