---
title: "Speaker Clustering in Textual Dialogue with Pairwise Utterance"
description: "The paper tackles a gap in dialogue data: many multi-party transcripts, such as subtitles, contain the… 44 references, cited by 1 paper here."
authors: ["Zhihua Su", "Qiang Zhou"]
year: 2022
venue: "Proceedings of the 29th International Conference on Computational Linguistics"
source: "https://aclanthology.org/2022.coling-1.61/"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/su-2022-speaker-clustering-textual-dialogue-pairwise/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/su-2022-speaker-clustering-textual-dialogue-pairwise/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2022](https://md.research.elamur.ai/free/en/papers/2022/page/6/) / Speaker Clustering in Textual Dialogue with Pairwise Utterance Relation and Cross-corpus Dialogue Act Supervision

# Speaker Clustering in Textual Dialogue with Pairwise Utterance Relation and Cross-corpus Dialogue Act Supervision

**Authors:** Zhihua Su and Qiang Zhou · **Year:** 2022 · **Venue:** [Proceedings of the 29th International Conference on Computational Linguistics](https://md.research.elamur.ai/free/en/venues/coling/) ([COLING 2022](https://md.research.elamur.ai/free/en/venues/coling/2022/)) · **Publisher:** [International Committee on Computational Linguistics (ICCL)](https://md.research.elamur.ai/free/en/publishers/international-committee-on-computational-linguistics-iccl/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2022.coling-1.61/) · [PDF](https://aclanthology.org/2022.coling-1.61.pdf) · [OpenAlex](https://openalex.org/W6891698868)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · 1 in our library, of the 141,840 papers whose bibliographies we read

## Summary
The paper tackles a gap in dialogue data: many multi-party transcripts, such as subtitles, contain the utterances but not who spoke them. The authors frame speaker recovery as clustering. Utterances are grouped so that each group belongs to one real speaker, with no preset list of candidates.

The method has four steps:
- A single BERT encodes the whole dialogue, with a [CLS] marker before every utterance.
- Each utterance vector combines mean and max pooling over its tokens.
- A bilinear scorer with a sigmoid predicts, for every utterance pair, whether the speaker is the same.
- Spectral clustering turns the resulting pairwise matrix into speaker labels.

Training adds an auxiliary dialogue act (DA) classification task. Its DA labels can come from other corpora, through a separate head for each dataset.

The model beats a sequence classification baseline. The DA supervision helps on almost all tests. Additional analyses cover three questions:
- how to estimate the number of speakers;
- whether the model confuses speaker identity with plain topical similarity;
- how prediction error grows with the distance between utterances.

The work matters because it could make unlabelled dialogue text usable for speaker-aware models.

## Key points
- **Input format.** The dialogue is fed as `[CLS] u1 [CLS] u2 … [CLS] uT [SEP]` into one BERT. Each utterance embedding (length 2·d_BERT) joins mean pooling and max pooling over its tokens.
- **Pairwise scoring.** Same-speaker probability is σ(c_mᵀ W c_n + b), computed only for pairs with m \< n. Binary cross-entropy is averaged over the T(T−1)/2 pairs.
- **Loss and positions.** The total loss is L = L_mat + λ·L_DA, with set-specific DA heads. Adding extra positional encodings gave no gain, because BERT's own positions were enough.
- **Number of speakers.** The cluster count comes from a spectral-gap rule: an eigenvalue threshold tuned on validation data. Results are reported both with this estimate and with the true speaker count.
- **Similarity control on MRDA.** The test used a balanced pair set and compared predictions against cosine similarity from an untuned BERT-base-uncased.
  - Predictions track real speaker relations strongly (t-test p = 4.49×10⁻¹⁰⁴).
  - Their link to semantic similarity (e.g. Spearman 0.0866) is about as weak as the inherent link between speakers and semantics.
- **Distance effect.** An error heatmap on the Ubuntu test set shows that adjacent pairs are modelled best. Distant pairs are harder, but their mean error still stays below 0.5.

## Interesting ideas
- Instead of labelling items one by one, predict a symmetric "same / different" matrix over all pairs and let spectral clustering derive consistent groups, including how many groups there are.
- An auxiliary task (dialogue acts) can be borrowed from a different corpus by giving each dataset its own classification head on a shared encoder.
- To check that a model learned the intended relation and not a shortcut, the authors run three correlation tests: prediction vs. truth, prediction vs. a confound (here, untuned-embedding similarity), and truth vs. confound.
- Encoding a whole dialogue in one pass, with a marker token per utterance, lets attention cross utterance boundaries at low cost.
- Plotting mean error by position pair (m, n) as a heatmap quickly shows where a pairwise model degrades with distance.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (frames): Coreference could score mention pairs and then cluster them into entity chains, keeping the pair scores as explainable evidence for each merge.
- **I2** (learn): Speech-act or role labels from heterogeneous treebanks and corpora could supervise one shared model without first unifying their tag sets.
- **I3** (learn): This is a ready-made negative-control template, e.g. testing whether coreference or answer decisions merely follow lexical overlap.
- **I4** (frames): Speech-act tagging in prag could encode the full exchange at once, so question–answer and offer–acceptance links shape each utterance's representation.
- **I5** (world): The same diagnostic could reveal at what textual distance event ordering or entity tracking in the world model starts to fail.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works

🧭 Topics (2): [🗨️ Dialogue systems](https://md.research.elamur.ai/free/en/topics/dialogue-systems/) · 401 works · [💬 Speech act theory](https://md.research.elamur.ai/free/en/topics/speech-act-theory/) · 160 works

## Cite

Cite as: Zhihua Su and Qiang Zhou. 2022. Speaker Clustering in Textual Dialogue with Pairwise Utterance Relation and Cross-corpus Dialogue Act Supervision. Proceedings of the 29th International Conference on Computational Linguistics.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{su2022speaker,
  title = {{Speaker Clustering in Textual Dialogue with Pairwise Utterance Relation and Cross-corpus Dialogue Act Supervision}},
  author = {Zhihua Su and Qiang Zhou},
  booktitle = {Proceedings of the 29th International Conference on Computational Linguistics},
  year = {2022},
  doi = {10.48448/fnfr-nd12},
  url = {https://doi.org/10.48448/fnfr-nd12}
}
```

- [← previous Siamese AraBERT-LSTM Model based Approach for Arabic Paraphrase Detection](https://md.research.elamur.ai/free/en/papers/mahmoud-2022-siamese-arabert-lstm-model-based/)
- [next → Study of Indian English Pronunciation Variabilities relative to Received Pronunciation](https://md.research.elamur.ai/free/en/papers/pal-2022-study-indian-english-pronunciation-variabilities/)
- [↑ back to papers of 2022](https://md.research.elamur.ai/free/en/papers/2022/page/6/)

## References

44 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Daniel Adiwardana, Minh-Thang Luong, David R. So, Jamie Hall, Noah Fiedel, Romal Thoppilan, Zi Yang, Apoorv Kulshreshtha, Gaurav Nemade, Yifeng Lu, and Quoc V. Le. 2020. [Towards a human-like open-domain chatbot](https://md.research.elamur.ai/free/en/papers/adiwardana-2020-towards-human-like-open-domain/). CoRR, abs/2001.09977. [↗](https://arxiv.org/abs/2001.09977)
2. Zakaria Aldeneh, Dimitrios Dimitriadis, and Emily Mower Provost. 2018. Improving end-of-turn detection in spoken dialogues by detecting speaker intentions as a secondary task. In 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 6159–6163. [find](https://scholar.google.com/scholar?q=Zakaria+Aldeneh%2C+Dimitrios+Dimitriadis%2C+and+Emily+Mower+Provost%2E+2018%2E+Improving+end%2Dof%2Dturn+detection+in+spoken+dialogues+by+detecting+speaker+intentions+as+a+secondary+task%2E+In+2018+IEEE+Internation)
3. Xavier Anguera, Simon Bozonnet, Nicholas Evans, Corinne Fredouille, Gerald Friedland, and Oriol Vinyals. 2012. Speaker diarization: A review of recent research. IEEE Transactions on Audio, Speech, and Language Processing, 20(2):356–370. [↗](https://doi.org/10.1109/tasl.2011.2125954)
4. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gray, Benjamin Chess, Jack Clark, Christopher Berner, Sam McCandlish, Alec Radford, Ilya Sutskever, and Dario Amodei. 2020. [Language models are few-shot learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/). In Advances in Neural Information Processing Systems, volume 33, pages 1877–1901. Curran Associates, Inc. We propose a text-based dialogue speaker clustering model. Based on the theory of the dialogue Harry Bunt, Jan Alexandersson, Jean Carletta, Jae-structure, the model holds the semantic content and Woong Choe, Alex Chengyu Fang, Koiti Hasida, the communicative functions explicitly with the Kiyong Lee, Volha Petukhova, Andrei Popescu-Belis, 742
5. Laurent Romary, Claudia Soria, and David Traum. 2010. [Towards an ISO Standard for Dialogue Act Annotation](https://md.research.elamur.ai/free/en/papers/bunt-2010-iso-standard-dialogue-act-annotation/). In Seventh conference on International Language Resources and Evaluation (LREC’10), La Valette, Malta.
6. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. [BERT: Pre-training of deep bidirectional transformers for language understanding](https://md.research.elamur.ai/free/en/papers/devlin-2019-bert-pre-training-deep-bidirectional/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 4171–4186, Minneapolis, Minnesota. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/n19-1423)
7. Adam Ek, Mats Wirén, Robert Östling, Kristina N. Björkenstam, Gintarė Grigonytė, and Sofia Gustafson Capková. 2018. Identifying speakers and addressees in dialogues extracted from literary fiction. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan. European Language Resources Association (ELRA). [find](https://scholar.google.com/scholar?q=Adam+Ek%2C+Mats+Wir%C3%A9n%2C+Robert+%C3%96stling%2C+Kristina+N%2E+Bj%C3%B6rkenstam%2C+Gintare%CC%87+Grigonyte%CC%87%2C+and+Sofia+Gustafson+Capkov%C3%A1%2E+2018%2E+Identifying+speakers+and+addressees+in+dialogues+extracted+from+literary+fiction%2E+)
8. Yihui Fu, Luyao Cheng, Shubo Lv, Yukai Jv, Yuxiang Kong, Zhuo Chen, Yanxin Hu, Lei Xie, Jian Wu, Hui Bu, Xin Xu, Jun Du, and Jingdong Chen. 2021. AISHELL-4: An Open Source Dataset for Speech Enhancement, Separation, Recognition and Speaker Diarization in Conference Scenario. In Proc. Interspeech 2021, pages 3665–3669. [find](https://scholar.google.com/scholar?q=Yihui+Fu%2C+Luyao+Cheng%2C+Shubo+Lv%2C+Yukai+Jv%2C+Yuxiang+Kong%2C+Zhuo+Chen%2C+Yanxin+Hu%2C+Lei+Xie%2C+Jian+Wu%2C+Hui+Bu%2C+Xin+Xu%2C+Jun+Du%2C+and+Jingdong+Chen%2E+2021%2E+AISHELL%2D4%3A+An+Open+Source+Dataset+for+Speech+Enhanceme)
9. Jia-Chen Gu, Chongyang Tao, Zhenhua Ling, Can Xu, Xiubo Geng, and Daxin Jiang. 2021. MPC-BERT: A pre-trained language model for multi-party conversation understanding. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 3682–3692, Online. Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.2106.01541)
10. Sibylle Hess, Wouter Duivesteijn, Philipp Honysz, and Katharina Morik. 2019. The spectacl of noncon-vex clustering: A spectral approach to density-based clustering. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):3788–3795. [find](https://scholar.google.com/scholar?q=Sibylle+Hess%2C+Wouter+Duivesteijn%2C+Philipp+Honysz%2C+and+Katharina+Morik%2E+2019%2E+The+spectacl+of+noncon%2Dvex+clustering%3A+A+spectral+approach+to+density%2Dbased+clustering%2E+Proceedings+of+the+AAAI+Conference+)
11. Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Veselin Stoyanov, and Luke Zettlemoyer. 2020. [BART: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension](https://md.research.elamur.ai/free/en/papers/lewis-2020-bart-denoising-sequence-sequence-pre/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7871–7880, Online. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/2020.acl-main.703)
12. Ruizhe Li, Chenghua Lin, Matthew Collinson, Xiao Li, and Guanyi Chen. 2019. A dual-attention hierarchical recurrent neural network for dialogue act classification. In Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL), pages 383–392, Hong Kong, China. Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.1810.09154)
13. Wei Li and Yunfang Wu. 2016. [Multi-level gated recurrent neural network for dialog act classification](https://md.research.elamur.ai/free/en/papers/li-2016-multi-level-gated-recurrent-neural/). In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 1970–1979, Osaka, Japan.
14. The COL-ING 2016 Organizing Committee. [find](https://scholar.google.com/scholar?q=The+COL%2DING+2016+Organizing+Committee%2E)
15. Yuhai Liang and Qiang Zhou. 2020. Detect turn-takings in subtitle streams with semantic recall transformer encoder. In 2020 International Conference on Asian Language Processing (IALP), pages 1–6. [find](https://scholar.google.com/scholar?q=Yuhai+Liang+and+Qiang+Zhou%2E+2020%2E+Detect+turn%2Dtakings+in+subtitle+streams+with+semantic+recall+transformer+encoder%2E+In+2020+International+Conference+on+Asian+Language+Processing+%28IALP%29%2C+pages+1%E2%80%936%2E)
16. Qingjian Lin, Ruiqing Yin, Ming Li, Hervé Bredin, and Claude Barras. 2019. LSTM Based Similarity Measurement with Spectral Clustering for Speaker Diarization. In Proc. Interspeech 2019, pages 366– 370. [find](https://scholar.google.com/scholar?q=Qingjian+Lin%2C+Ruiqing+Yin%2C+Ming+Li%2C+Herv%C3%A9+Bredin%2C+and+Claude+Barras%2E+2019%2E+LSTM+Based+Similarity+Measurement+with+Spectral+Clustering+for+Speaker+Diarization%2E+In+Proc%2E+Interspeech+2019%2C+pages+366%E2%80%93+370)
17. Pierre Lison, Jörg Tiedemann, and Milen Kouylekov. 2018. [OpenSubtitles2018: Statistical rescoring of sentence alignments in large, noisy parallel corpora](https://md.research.elamur.ai/free/en/papers/lison-2018-opensubtitles2018-statistical-rescoring-sentence-alignments/). In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan. European Language Resources Association (ELRA).
18. Lawrence Hubert and Phipps Arabie. 1985. Comparing partitions. Journal of classification, 2(1):193–218. [find](https://scholar.google.com/scholar?q=Lawrence+Hubert+and+Phipps+Arabie%2E+1985%2E+Comparing+partitions%2E+Journal+of+classification%2C+2%281%29%3A193%E2%80%93218%2E)
19. Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, May 6-9, 2019. OpenRe-view.net. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoupled+weight+decay+regularization%2E+In+7th+International+Conference+on+Learning+Representations%2C+ICLR+2019%2C+New+Orleans%2C+LA%2C+USA%2C+May+6%2D9%2C+2019%2E+OpenRe%2Dview%2E)
20. Naoyuki Kanda, Xiong Xiao, Yashesh Gaur, Xiaofei Wang, Zhong Meng, Zhuo Chen, and Takuya Yoshioka. 2022. Transcribe-to-diarize: Neural speaker diarization for unlimited number of speakers using end-to-end speaker-attributed asr. In ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 8082– 8086. [↗](https://doi.org/10.48550/arxiv.2110.03151)
21. Ryan Lowe, Nissan Pow, Iulian Serban, and Joelle Pineau. 2015. [The Ubuntu dialogue corpus: A large dataset for research in unstructured multi-turn dialogue systems](https://md.research.elamur.ai/free/en/papers/lowe-2015-ubuntu-dialogue-corpus-large-dataset/). In Proceedings of the 16th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 285–294, Prague, Czech Republic. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/w15-4640)
22. Amitava Kundu, Dipankar Das, and Sivaji Bandyopadhyay. 2012. Speaker identification from film dialogues. In 2012 4th International Conference on Intelligent Human Computer Interaction (IHCI), pages 1–4. [find](https://scholar.google.com/scholar?q=Amitava+Kundu%2C+Dipankar+Das%2C+and+Sivaji+Bandyopadhyay%2E+2012%2E+Speaker+identification+from+film+dialogues%2E+In+2012+4th+International+Conference+on+Intelligent+Human+Computer+Interaction+%28IHCI%29%2C+pages+1%E2%80%93)
23. Yanick Lukic, Carlo Vogt, Oliver Dürr, and Thilo Stadel-mann. 2016. Speaker identification and clustering using convolutional neural networks. In 2016 IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP), pages 1–6. [find](https://scholar.google.com/scholar?q=Yanick+Lukic%2C+Carlo+Vogt%2C+Oliver+D%C3%BCrr%2C+and+Thilo+Stadel%2Dmann%2E+2016%2E+Speaker+identification+and+clustering+using+convolutional+neural+networks%2E+In+2016+IEEE+26th+International+Workshop+on+Machine+Learn)
24. Kaixin Ma, Catherine Xiao, and Jinho D. Choi. 2017. Text-based speaker identification on multiparty dialogues using multi-document convolutional neural networks. In Proceedings of ACL 2017, Student Research Workshop, pages 49–55, Vancouver, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Kaixin+Ma%2C+Catherine+Xiao%2C+and+Jinho+D%2E+Choi%2E+2017%2E+Text%2Dbased+speaker+identification+on+multiparty+dialogues+using+multi%2Ddocument+convolutional+neural+networks%2E+In+Proceedings+of+ACL+2017%2C+Student+Re)
25. Xiaofei Ma, Zhiguo Wang, Patrick Ng, Ramesh Nal-lapati, and Bing Xiang. 2019. Universal text representation from BERT: an empirical study. CoRR, abs/1910.07973. [↗](https://arxiv.org/abs/1910.07973)
26. Kai Nickel, Tobias Gehrig, Rainer Stiefelhagen, and John McDonough. 2005. A joint particle filter for audio-visual speaker tracking. In Proceedings of the 7th International Conference on Multimodal Interfaces, ICMI ’05, page 61–68, New York, NY, USA. Association for Computing Machinery. [find](https://scholar.google.com/scholar?q=Kai+Nickel%2C+Tobias+Gehrig%2C+Rainer+Stiefelhagen%2C+and+John+McDonough%2E+2005%2E+A+joint+particle+filter+for+audio%2Dvisual+speaker+tracking%2E+In+Proceedings+of+the+7th+International+Conference+on+Multimodal+In)
27. Hiroki Ouchi and Yuta Tsuboi. 2016. Addressee and response selection for multi-party conversation. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2133– 2143, Austin, Texas. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Hiroki+Ouchi+and+Yuta+Tsuboi%2E+2016%2E+Addressee+and+response+selection+for+multi%2Dparty+conversation%2E+In+Proceedings+of+the+2016+Conference+on+Empirical+Methods+in+Natural+Language+Processing%2C+pages+2133)
28. Tae Jin Park, Naoyuki Kanda, Dimitrios Dimitriadis, Kyu J. Han, Shinji Watanabe, and Shrikanth Narayanan. 2022. A review of speaker diarization: Recent advances with deep learning. Computer Speech & Language, 72:101317. [↗](https://doi.org/10.1016/j.csl.2021.101317)
29. Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, Alban Desmaison, Andreas Kopf, Edward Yang, Zachary DeVito, Martin Raison, Alykhan Te-jani, Sasank Chilamkurthy, Benoit Steiner, Lu Fang, Junjie Bai, and Soumith Chintala. 2019. [Pytorch: An imperative style, high-performance deep learning library](https://md.research.elamur.ai/free/en/papers/paszke-2019-pytorch-imperative-style-high-performance/). In Advances in Neural Information Processing Systems, volume 32. Curran Associates, Inc. [↗](https://doi.org/10.48550/arxiv.1912.01703)
30. Desh Raj, Pavel Denisov, Zhuo Chen, Hakan Erdogan, Zili Huang, Maokui He, Shinji Watanabe, Jun Du, Takuya Yoshioka, Yi Luo, Naoyuki Kanda, Jinyu Li, Scott Wisdom, and John R. Hershey. 2021. Integration of speech separation, diarization, and recognition for multi-speaker meetings: System description, comparison, and analysis. In 2021 IEEE Spoken Language Technology Workshop (SLT), pages 897–904. [↗](https://doi.org/10.17023/994t-7n12)
31. Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Eric Michael Smith, Y-Lan Boureau, and Jason Weston. 2021. [Recipes for building an open-domain chatbot](https://md.research.elamur.ai/free/en/papers/roller-2021-recipes-building-open-domain-chatbot/). In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 300–325, Online. Association for Computational Linguistics.
32. Emanuel A. Schegloff and Harvey Sacks. 1973. Opening up closings. Semiotica, 8(4). [↗](https://doi.org/10.1515/semi.1973.8.4.289)
33. John R Searle and John Rogers Searle. 1969. Speech acts: An essay in the philosophy of language, volume 626. Cambridge university press. [find](https://scholar.google.com/scholar?q=John+R+Searle+and+John+Rogers+Searle%2E+1969%2E+Speech+acts%3A+An+essay+in+the+philosophy+of+language%2C+volume+626%2E+Cambridge+university+press%2E)
34. Elizabeth Shriberg, Raj Dhillon, Sonali Bhagat, Jeremy Ang, and Hannah Carvey. 2004. [The ICSI meeting recorder dialog act (MRDA) corpus](https://md.research.elamur.ai/free/en/papers/shriberg-2004-icsi-meeting-recorder-dialog-act/). In Proceedings of the 5th SIGdial Workshop on Discourse and Dialogue at HLT-NAACL 2004, pages 97–100, Cambridge, Massachusetts, USA. Association for Computational Linguistics. [↗](https://doi.org/10.21236/ada460980)
35. Andreas Stolcke, Klaus Ries, Noah Coccaro, Elizabeth Shriberg, Rebecca Bates, Daniel Jurafsky, Paul Taylor, Rachel Martin, Carol Van Ess-Dykema, and Marie Meteer. 2000. [Dialogue act modeling for automatic tagging and recognition of conversational speech](https://md.research.elamur.ai/free/en/papers/stolcke-2000-dialogue-act-modeling-automatic-tagging/). Computational Linguistics, 26(3):339–374. [↗](https://doi.org/10.1162/089120100561737)
36. Dinoj Surendran and Gina-Anne Levow. 2006. Dialog act tagging with support vector machines and hidden markov models. In Ninth International Conference on Spoken Language Processing. [find](https://scholar.google.com/scholar?q=Dinoj+Surendran+and+Gina%2DAnne+Levow%2E+2006%2E+Dialog+act+tagging+with+support+vector+machines+and+hidden+markov+models%2E+In+Ninth+International+Conference+on+Spoken+Language+Processing%2E)
37. S.E. Tranter and D.A. Reynolds. 2006. An overview of automatic speaker diarization systems. IEEE Transactions on Audio, Speech, and Language Processing, 14(5):1557–1565. [↗](https://doi.org/10.1109/tasl.2006.878256)
38. Iulia Turc, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. [Well-read students learn better: The impact of student initialization on knowledge distillation](https://md.research.elamur.ai/free/en/papers/turc-2019-well-read-students-learn-better/). CoRR, abs/1908.08962. [↗](https://arxiv.org/abs/1908.08962)
39. Félicien Vallet, Jim Uro, Jérémy Andriamakaoly, Hakim Nabi, Mathieu Derval, and Jean Carrive. 2016. Speech trax: A bottom to the top approach for speaker tracking and indexing in an archiving context. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), pages 2011–2016, Portorož, Slovenia. European Language Resources Association (ELRA). [find](https://scholar.google.com/scholar?q=F%C3%A9licien+Vallet%2C+Jim+Uro%2C+J%C3%A9r%C3%A9my+Andriamakaoly%2C+Hakim+Nabi%2C+Mathieu+Derval%2C+and+Jean+Carrive%2E+2016%2E+Speech+trax%3A+A+bottom+to+the+top+approach+for+speaker+tracking+and+indexing+in+an+archiving+context%2E)
40. Ulrike Von Luxburg. 2007. A tutorial on spectral clustering. Statistics and computing, 17(4):395–416. [↗](https://doi.org/10.1007/s11222-007-9033-z)
41. Ce Wang and M.S. Brandstein. 1999. Multi-source face tracking with audio and visual data. In 1999 IEEE Third Workshop on Multimedia Signal Processing (Cat. No.99TH8451), pages 169–174. [find](https://scholar.google.com/scholar?q=Ce+Wang+and+M%2ES%2E+Brandstein%2E+1999%2E+Multi%2Dsource+face+tracking+with+audio+and+visual+data%2E+In+1999+IEEE+Third+Workshop+on+Multimedia+Signal+Processing+%28Cat%2E+No%2E99TH8451%29%2C+pages+169%E2%80%93174%2E)
42. Han Xiao. 2018. bert-as-service. [https://github.com/hanxiao/bert-as-service](https://github.com/hanxiao/bert-as-service). [↗](https://github.com/hanxiao/bert-as-service)
43. Xuejing Zhang, Xueqiang Lv, and Qiang Zhou. 2018. Chinese dialogue analysis using multi-task learning framework. In 2018 International Conference on Asian Language Processing (IALP), pages 102–107. [find](https://scholar.google.com/scholar?q=Xuejing+Zhang%2C+Xueqiang+Lv%2C+and+Qiang+Zhou%2E+2018%2E+Chinese+dialogue+analysis+using+multi%2Dtask+learning+framework%2E+In+2018+International+Conference+on+Asian+Language+Processing+%28IALP%29%2C+pages+102%E2%80%93107%2E)
44. Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, and Bill Dolan. 2020. [DIALOGPT : Large-scale generative pre-training for conversational response generation](https://md.research.elamur.ai/free/en/papers/zhang-2019-dialogpt-large-scale-generative-pre/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 270–278, Online. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/2020.acl-demos.30)

*Updated 24 October 2026*