Neural Extractive Text Summarization with Syntactic Compression
Preprint
Abstract
Recent neural network approaches to summarization are largely either selection-based extraction or generation-based abstraction. In this work, we present a neural model for single-document summarization based on joint extraction and syntactic compression. Our model chooses sentences from the document, identifies possible compressions based on constituency parses, and scores those compressions with a neural model to produce the final summary. For learning, we construct oracle extractive-compressive summaries, then learn both of our components jointly with this supervision. Experimental results on the CNN/Daily Mail and New York Times datasets show that our model achieves strong performance (comparable to state-of-the-art systems) as evaluated by ROUGE. Moreover, our approach outperforms an off-the-shelf compression module, and human and manual evaluation shows that our model's output generally remains grammatical.Keywords
All Related Versions
This publication has 19 references indexed in Scilit:
- Newsroom: A Dataset of 1.3 Million Summaries with Diverse Extractive Strategies2018
- Abstractive Document Summarization with a Graph-Based Attentional Neural Model2017
- Language as a Latent Variable: Discrete Generative Models for Sentence Compression2016
- Abstractive Text Summarization using Sequence-to-sequence RNNs and Beyond2016
- Abstractive Sentence Summarization with Attentive Recurrent Neural Networks2016
- Sentence Compression by Deletion with LSTMs2015
- A Neural Attention Model for Abstractive Sentence Summarization2015
- Sentence Compression as Tree TransductionJournal of Artificial Intelligence Research, 2009
- Summarization beyond sentence extraction: A probabilistic approach to sentence compressionArtificial Intelligence, 2002
- The use of MMR, diversity-based reranking for reordering documents and producing summaries1998