Fact-level Extractive Summarization with Hierarchical Graph Mask on BERT

Abstract

Most current extractive summarization models generate summaries by selecting salient sentences. However, one of the problems with sentence-level extractive summarization is that there exists a gap between the human-written gold summary and the oracle sentence labels. In this paper, we propose to extract fact-level semantic units for better extractive summarization. We also introduce a hierarchical structure, which incorporates the multi-level of granularities of the textual information into the model. In addition, we incorporate our model with BERT using a hierarchical graph mask. This allows us to combine BERT's ability in natural language understanding and the structural information without increasing the scale of the model. Experiments on the CNN/DaliyMail dataset show that our model achieves state-of-the-art results.

arXiv commentAccept by Coling2020

References in corpus (14)

• Teaching Machines to Read and Comprehend
• A Deep Reinforced Model for Abstractive Summarization
• SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents
• Faithful to the Original: Fact Aware Neural Abstractive Summarization
• Text Summarization with Pretrained Encoders
• Neural Summarization by Extracting Sentences and Words
• HIBERT: Document Level Pre-training of Hierarchical Bidirectional Transformers for Document Summarization
• Text Generation from Knowledge Graphs with Graph Transformers
• Hierarchical Graph Network for Multi-hop Question Answering
• BanditSum: Extractive Summarization as a Contextual Bandit
• Searching for Effective Neural Extractive Summarization: What Works and What's Next
• Neural Document Summarization by Jointly Learning to Score and Select Sentences
• Neural Latent Extractive Document Summarization
• Summary Level Training of Sentence Rewriting for Abstractive Summarization