Abstractive Dialogue Summarization with Sentence-Gated Modeling Optimized by Dialogue Acts
This work addresses the challenge of summarizing multi-speaker dialogues, which is incremental as it builds on prior single-speaker summarization methods by integrating interactive signals.
The paper tackled the problem of abstractive summarization for dialogues by explicitly incorporating dialogue acts into a neural model, resulting in significant performance improvements over state-of-the-art baselines on the AMI meeting corpus.
Neural abstractive summarization has been increasingly studied, where the prior work mainly focused on summarizing single-speaker documents (news, scientific publications, etc). In dialogues, there are different interactions between speakers, which are usually defined as dialogue acts. The interactive signals may provide informative cues for better summarizing dialogues. This paper proposes to explicitly leverage dialogue acts in a neural summarization model, where a sentence-gated mechanism is designed for modeling the relationship between dialogue acts and the summary. The experiments show that our proposed model significantly improves the abstractive summarization performance compared to the state-of-the-art baselines on AMI meeting corpus, demonstrating the usefulness of the interactive signal provided by dialogue acts.