From spoken dialogue to formal summary: An utterance rewriting for dialogue summarization

Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 3859 - 3869 July 10-15, 2022 ©2022 Association for Computational Linguistics From spoken dialogue to formal summary: An utterance rewriting for dialogue summarization Yue Fang 1∗, Hainan Zhang 2, Hongshen Chen 2, Zhuoye Ding 2, Bo Long 2, Yanyan Lan 2 and Yanquan Zhou 1 1Beijing University of Posts and Telecommunications, Beijing, China 2JD.com, Beijing, China {fangyue, zhouyanquan}@bupt.edu.com, zhanghainan1990@163.com,ac@chenhongshen.com, lanyanyan@ict.ac.cn,{dingzhuoye, bo.long}@jd.com Abstract Due to the dialogue characteristics of unstruc- tured contexts and multi-parties with first- person perspective, many successful text sum- marization works have failed when dealing with dialogue summarization. In dialogue sum- marization task, the input dialogue is usually spoken style with ellipsis and co-references but the output summaries are more formal and com- plete. Therefore, the dialogue summarization model should be able to complete the ellipsis content and co-reference information and then produce a suitable summary accordingly. How- ever, the current state-of-the-art models pay more attention on the topic or structure of sum- mary, rather than the consistency of dialogue summary with its input dialogue context, which may suffer from the personal and logical incon- sistency problem. In this paper, we propose a new model, named ReWriteSum, to tackle this problem. Firstly, an utterance rewriter is conducted to complete the ellipsis content of dialogue content and then obtain the rewriting utterances. Then, the co-reference data aug- mentation mechanism is utilized to replace the referential person name with its specific name to enhance the personal information. Finally, the rewriting utterances and the co-reference re- placement data are used in the standard BART model. Experimental results on both SAMSum and DialSum datasets show that our ReWrite- Sum significantly outperforms baseline models, in terms of both metric-based and human evalu- ations. Further analysis on multi-speakers also shows that ReWriteSum can obtain relatively higher improvement with more speakers, vali- dating the correctness and property of ReWrite- Sum. 1 Introduction Despite many existing text summarization works on single-speaker written documents, such as news and encyclopedia articles (Rush et al., 2015; ∗Work done during internship at JD.com. Example one Ann: Hi, is the laptop still available? Josh: Yes it is. Ann: I can pay 200 dollars. Josh: The price is 250 and it’s non-negotiable. Ann: Do you have a bag for it? Some other accessories? Josh: I have a bag and a small usb mouse. Ann: Sounds good, I’ll take it, where can I pick it up? Ground-Truth: Ann wants to buy Josh’s laptop for $200. Josh doesn’t want to negotiate the price. Ann will take it for $250 with accessories. BART Prediction: Ann will pay 200 dollars for the laptop and the price is non-negotiable. Ann will pick a bag from Josh. Example two Mike: Dude, Wendy has grown prettier. Dave: I know right? Mike: Yeah, since she came from Houston, she looks like an angel. Dave: I’ll have to hit on her soon. Mike: Haha, stay off, I hear Jerry is her …

Content truncated to 500 words