In this research work, we present a method to generate summaries of long scientific documents that uses the advantages of both extractive and abstractive approaches. Before producing a summary in an abstractive manner, we perform the extractive step, which then is used for conditioning the abstractor module. We used pre-trained transformer-based language models, for both extractor and abstractor. Our experiments showed that using extractive and abstractive models jointly significantly improves summarization results and ROUGE scores.
@article{arxiv.2006.05354,
title = {Combination of abstractive and extractive approaches for summarization of long scientific texts},
author = {Vladislav Tretyak and Denis Stepanov},
journal= {arXiv preprint arXiv:2006.05354},
year = {2020}
}
Comments
11 pages, 2 figures, 3 table, submitted to 23rd International Conference on Discovery Science. Fixed authors list