Structure-Aware Abstractive Conversation Summarization via Discourse and Action Graphs

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1380–1391 June 6–11, 2021. ©2021 Association for Computational Linguistics 1380 Structure-Aware Abstractive Conversation Summarization via Discourse and Action Graphs Jiaao Chen School of Interactive Computing Georgia Institute of Technology jiaaochen@gatech.edu Diyi Yang School of Interactive Computing Georgia Institute of Technology dyang888@gatech.edu Abstract Abstractive conversation summarization has received much attention recently. However, these generated summaries often suffer from insufficient, redundant, or incorrect content, largely due to the unstructured and complex characteristics of human-human interactions. To this end, we propose to explicitly model the rich structures in conversations for more precise and accurate conversation summariza- tion, by first incorporating discourse relations between utterances and action triples (“WHO- DOING-WHAT”) in utterances through struc- tured graphs to better encode conversations, and then designing a multi-granularity decoder to generate summaries by combining all lev- els of information. Experiments show that our proposed models outperform state-of-the- art methods and generalize well in other do- mains in terms of both automatic evaluations and human judgments. We have publicly re- leased our code at https://github.com/ GT-SALT/Structure-Aware-BART. 1 Introduction Online interaction has become an indispensable component of everyday life and people are increas- ingly using textual conversations to exchange ideas, make plans, and share information. However, it is time-consuming to recap and grasp all the core con- tent within every complex conversation (Gao et al., 2020; Feng et al., 2020). As a result, how to or- ganize massive everyday interactions into natural, concise, and informative text, i.e., abstractive con- versation summarization, starts to gain importance. Significant progress has been made on abstrac- tive summarization for structured document via pointer generator (See et al., 2017), reinforcement methods (Paulus et al., 2018; Huang et al., 2020a) and pre-trained models (Liu and Lapata, 2019; Lewis et al., 2020; Zhang et al., 2019). Despite the huge success, it is challenging to directly apply doc- ument models to summarize conversations, due to Figure 1: An example of discourse relation graph (a) and action graph (b) from one conversation in SAM- Sum (Gliwa et al., 2019). The annotated summary is Simon was on the phone before, so he didn’t here He- len calling. Simon will fetch Helen some tissues. a set of inherent differences between conversations and documents (Gliwa et al., 2019). First, speaker interruptions like repetitions, false-starts, and hesi- tations are frequent in conversations (Sacks et al., 1978), and key information resides in different por- tions of a conversation. These unstructured proper- ties pose challenges for models to focus on salient contents that are necessary for generating both ab- stractive and informative summaries. Second, there is more than one speaker in conversations and peo- ple interact with each other in different language styles (Zhu et al., 2020b). The complex interactions among multiple speakers make it harder for mod- 1381 els to identify and associate speakers with correct actions so as to generate factual summaries. In order to summarize the unstructured and com- plex conversations, …

Content truncated to 500 words