/EACL2023_Tutorial_Dialogue_Summarization

EACL2023_Tutorial_Dialogue_Summarization

Materials for EACL2023 tutorial: Summarization of Dialogues and Conversations At Scale

Time and Location

  1. Time: 14:15 - 18:00 (CEST), May 5, 2023.

  2. Location: Online

  3. Live Stream on Zoom: [Link to be announced]

If you find this tutorial helpful for your work, please kindly cite our paper.

@inproceedings{yang2023summarization,
  title={Summarization of Dialogues and Conversations At Scale},
  author={Yang, Diyi and Zhu, Chenguang},
  booktitle={Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts},
  pages={13--18},
  year={2023}
}

Tutorial Abstract

Conversations are the natural communication format for people. This fact has motivated the large body of question answering and chatbot research as a seamless way for people to interact with machines. The conversations between people however, captured as video, audio or private or public written conversations, largely remain untapped as a source of compelling starting point for developing language technology. Summarizing such conversations can be enormously beneficial: automatic minutes for meetings or meeting highlights sent to relevant people can optimize communication in various groups while minimizing demands on people’s time; similarly analysis of conversations in online support groups can provide valuable information to doctors about the patient concerns. Summarizing written and spoken conversation poses unique research challenges—text reformulation, discourse and meaning analysis beyond the sentence, collecting data, and proper evaluation metrics. All these have been revisited by researchers since the emergence of neural approaches as the dominant approach for solving language processing problems. In this tutorial, we will survey the cutting-edge methods for summarization of conversations, covering key sub-areas whose combination is needed for a successful solution.

Tutorial Materials

1. Slides

[Tutorial Slides]

2. Video

[Tutorial Video] (Need EACL registration at Underline)

3. Survey

  • A Survey on Dialogue Summarization: Recent Advances and New Frontiers, in IJCAI 2021. [PDF]

4. Reading list:

  • DialogLM: Pre-trained Model for Long Dialogue Understanding and Summarization. M. Zhong et al. AAAI 2022. [pdf]
  • A Hierarchical Network for Abstractive Meeting Summarization with Cross-Domain Pretraining. Findings of EMNLP 2020. [pdf]
  • QMSum: A New Benchmark for Query-based Multi-domain Meeting Summarization. M. Zhong et al. NAACL 2021. [pdf]
  • A Sliding-Window Approach to Automatic Creation of Meeting Minutes. J. J. Koay et al. NAACL: Student Research Workshop, 2021. [pdf]
  • Improving Abstractive Dialogue Summarization with Speaker-Aware Supervised Contrastive Learning. Z. Geng et al. COLING 2022. [pdf]
  • Coreference-Aware Dialogue Summarization. Z. Liu et al. SIGDIAL 2021. [pdf]
  • Improving Abstractive Dialogue Summarization with Graph Structures and Topic Words. L. Zhao et al. COLING 2020. [pdf]
  • How Domain Terminology Affects Meeting Summarization Performance. J. J. Koay et al. COLING 2020. [pdf]
  • Dr. Summarize: Global Summarization of Medical Dialogue by Exploiting Local Structures. A. Joshi et al. Findings of EMNLP 2020. [pdf]
  • Longformer: The Long-Document Transformer. I. Beltagy et al. arXiv 2020. [pdf]

Tutorial schedule (Tentative)

Local time (CEST) Content Presenter
14:15-14:45 Introduction to Conversation Summarization Diyi Yang
14:45-15:15 Conversation Structures and Evaluation Diyi Yang
15:15-15:30 Break -
15:30-16:50 Pretraining and Models Chenguang Zhu
16:50-17:00 Conclusion and Future Directions Chenguang Zhu

Presenters

       

          Diyi Yang             Chenguang Zhu

Diyi Yang is an assistant professor in the Computer Science Department at Stanford University. Her research focuses on dialogue summarization, learning with limited and noisy text data, user-centric language generation, and computational social science. Diyi has organized four workshops at NLP conferences: Widening NLP Workshops at NAACL 2018 and ACL 2019, Casual Inference workshop at EMNLP 2021, and NLG Evaluation workshop at EMNLP 2021. She also gave a tutorial at ACL 2022 on Learning with Limited Data.

Chenguang Zhu is a Principal Research Manager in Microsoft Cognitive Services Research Group, where he leads the Knowledge & Language Team. His research in NLP covers dialogue summarization, knowledge graph, prompt learning and multimodal learning. He has led teams to achieve first places in multiple NLP competitions, including CommonsenseQA, CommonGen, FEVER, CoQA, ARC and SQuAD v1.0. He holds a Ph.D. degree in Computer Science from Stanford University. Dr. Zhu has held Tutorial on Knowledge-Augmented NLP Methods at ACL 2022 and WSDM 2023, as well as the Workshop of Knowledge-Augmented NLP Methods at AAAI 2023.