We propose a unified model combining the strength of extractive and abstractive summarization. On the one hand, a simple extractive model can obtain sentence-level attention with high ROUGE scores but less readable. On the other hand, a more complicated abstractive model can obtain word-level dynamic attention to generate a more readable paragraph. In our model, sentence-level attention is used to modulate the word-level attention such that words in less attended sentences are less likely to be generated. Moreover, a novel inconsistency loss function is introduced to penalize the inconsistency between two levels of attentions. By end-to-end training our model with the inconsistency loss and original losses of extractive and abstractive models, we achieve state-of-the-art ROUGE scores while being the most informative and readable summarization on the CNN/Daily Mail dataset in a solid human evaluation.
@article{arxiv.1805.06266,
title = {A Unified Model for Extractive and Abstractive Summarization using Inconsistency Loss},
author = {Wan-Ting Hsu and Chieh-Kai Lin and Ming-Ying Lee and Kerui Min and Jing Tang and Min Sun},
journal= {arXiv preprint arXiv:1805.06266},
year = {2018}
}