---
title: "Dialogue Coherence Assessment Without Explicit Dialogue Act"
description: "Automatic dialogue evaluation needs a way to judge coherence. 3.2 points. 34 references, cited by 16 papers here."
authors: ["Mohsen Mesgar", "Sebastian Bücker", "Iryna Gurevych"]
year: 2020
venue: "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics"
source: "https://aclanthology.org/2020.acl-main.133/"
doi: "10.18653/v1/2020.acl-main.133"
arxiv: "1908.08486"
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/mesgar-2020-dialogue-coherence-assessment-without-explicit/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/mesgar-2020-dialogue-coherence-assessment-without-explicit/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2020](https://md.research.elamur.ai/free/en/papers/2020/page/6/) / Dialogue Coherence Assessment Without Explicit Dialogue Act Labels

# Dialogue Coherence Assessment Without Explicit Dialogue Act Labels

**Authors:** [Mohsen Mesgar](https://md.research.elamur.ai/free/en/authors/mohsen-mesgar/), Sebastian Bücker and [Iryna Gurevych](https://md.research.elamur.ai/free/en/authors/iryna-gurevych/) · **Year:** 2020 · **Venue:** [Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2020](https://md.research.elamur.ai/free/en/venues/acl/2020/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2020.acl-main.133/) · [PDF](https://aclanthology.org/2020.acl-main.133.pdf) · [DOI 10.18653/v1/2020.acl-main.133](https://doi.org/10.18653/v1/2020.acl-main.133) · [arXiv 1908.08486](https://arxiv.org/abs/1908.08486) · [OpenAlex](https://openalex.org/W3033475175)

**Citations:** 24 citations (OpenAlex, 2026-10-14) · [16](https://md.research.elamur.ai/free/en/papers/mesgar-2020-dialogue-coherence-assessment-without-explicit/cited-by/) in our library, of the 141,840 papers whose bibliographies we read

## Summary
Automatic dialogue evaluation needs a way to judge coherence. Word-overlap scores such as BLEU and costly human ratings both fall short. Earlier dialogue coherence models reused monologue features such as entity transitions and then added dialogue act (DA) labels on top. This has two costs: utterances are reduced to entity mentions, which are sparse in short turns, and accuracy depends on whoever supplies the DA labels.

The authors train a neural coherence scorer, DiCoh, together with a Dialogue Act Prediction (DAP) head. Both share one utterance encoder built from word embeddings, a BiLSTM and self-attention. DA labels serve only as a training signal, so the model needs no labels at test time. Coherence is learned by pairwise ranking: an original dialogue should score higher than a corrupted copy of itself.

On DailyDialog the model beats the strongest prior system by more than 20 accuracy points, and on SwitchBoard it is roughly on par. It also stays more stable when the training and test corruptions differ. This makes it a practical, label-free tool for comparing dialogue agents.

## Key points
- Evaluation uses DailyDialog and SwitchBoard with four corruption types: shuffling the order of utterances (UO), inserting an utterance (UI), swapping in a different utterance (UR), and shuffling only the even-numbered turns (EUO). The last two are new and dialogue-specific. The paper's example corruption reshuffles only one speaker's turns.
- Architecture:
  - Utterance vectors pass through a dialogue-level BiLSTM, then attention, then a linear layer that outputs a coherence score.
  - Training uses a margin hinge loss on dialogue pairs plus cross-entropy for DAP.
  - The two losses are combined with trainable balancing weights, following Kendall et al. (2018).
- In cross-perturbation tests on DailyDialog, M-DiCoh beat EAGrid on every test corruption. EAGrid's UO accuracy fell from 71.72% to 58.7% when it was trained on UR, a drop of about 13 points. M-DiCoh's largest drop on UO was 3.2 points.
- On DailyDialog, DAP F1 is around 77–79. The jointly trained M-DAP is on par with the standalone S-DAP, and both beat an SVM-BoW baseline. So DA supervision helps coherence, but coherence supervision does not help DA prediction.
- Code is released at github.com/UKPLab/acl2020-dialogue-coherence-assessment.

## Interesting ideas
- Use a costly label as an auxiliary training objective on a shared encoder instead of as a pipeline input, so inference never depends on a brittle upstream classifier.
- Create free supervision by corrupting a coherent text (reordering, inserting or replacing units) and training a model to rank the original above the corruption.
- Corruptions that touch only one speaker's turns stay locally plausible, so they make harder negatives than a full shuffle, which the paper found easiest to detect.
- Train on one corruption type and test on others to show whether a model learned coherence or just the quirks of one corruption.
- Let trainable weights balance several losses instead of hand-tuning their mix.

## For Mova

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

- **I1** (learn): Speech-act or semantic-role tags from silver data could act as auxiliary heads when training gparse, so the parser absorbs that signal while runtime does not wait on a separate classifier.
- **I2** (learn): Shuffled or spliced fables could serve as negative controls, checking that the world-of-text builder reports more contradictions in event order for the corrupted version.
- **I3** (world): Swapping only the events of one character in a tale would test whether the event-order and state tracking catches subtle inconsistencies, not just obvious chaos.
- **I4** (learn): Self-induced repair rules learned from one error type could be scored on other error types to expose rules that overfit their source.
- **I5** (tree): gparse's joint objectives (arcs, labels, tags) could be weighted this way, removing a manual hyperparameter search when treebanks or UD dialects change.

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

## Institutions

- [Technische Universität Darmstadt](https://md.research.elamur.ai/free/en/universities/technische-universitat-darmstadt/) — University, Darmstadt, Germany · 76 works · 35,008 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,085 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,888 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: Mohsen Mesgar, Sebastian Bücker and Iryna Gurevych. 2020. Dialogue Coherence Assessment Without Explicit Dialogue Act Labels. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.

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

```bibtex
@inproceedings{mesgar2020dialogue,
  title = {{Dialogue Coherence Assessment Without Explicit Dialogue Act Labels}},
  author = {Mohsen Mesgar, Sebastian Bücker and Iryna Gurevych},
  booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
  year = {2020},
  doi = {10.18653/v1/2020.acl-main.133},
  eprint = {1908.08486},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.18653/v1/2020.acl-main.133}
}
```

- [← previous TransS-Driven Joint Learning Architecture for Implicit Discourse Relation Recognition](https://md.research.elamur.ai/free/en/papers/he-2020-transs-driven-joint-learning-architecture/)
- [next → QANom: Question-Answer driven SRL for Nominalizations](https://md.research.elamur.ai/free/en/papers/klein-2020-qanom-question-answer-driven-srl/)
- [↑ back to papers of 2020](https://md.research.elamur.ai/free/en/papers/2020/page/6/)

## References

34 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Regina Barzilay and Mirella Lapata. 2005. Modeling local coherence: An entity-based approach. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics, Ann Arbor, Mich., 25–30 June 2005, pages 141–148. [find](https://scholar.google.com/scholar?q=Regina+Barzilay+and+Mirella+Lapata%2E+2005%2E+Modeling+local+coherence%3A+An+entity%2Dbased+approach%2E+In+Proceedings+of+the+43rd+Annual+Meeting+of+the+Association+for+Computational+Linguistics%2C+Ann+Arbor%2C+Mic)
2. Regina Barzilay and Mirella Lapata. 2008. [Modeling local coherence: An entity-based approach](https://md.research.elamur.ai/free/en/papers/barzilay-2008-modeling-local-coherence-entity-based/). Computational Linguistics, 34(1):1–34. [↗](https://doi.org/10.1162/coli.2008.34.1.1)
3. Jill Burstein, Joel Tetreault, and Slava Andreyev. 2010. Using entity-based features to model coherence in student essays. In Proceedings of Human Language Technologies 2010: The Conference of the North American Chapter of the Association for Computational Linguistics, Los Angeles, Cal., 2–4 June 2010, pages 681–684. [find](https://scholar.google.com/scholar?q=Jill+Burstein%2C+Joel+Tetreault%2C+and+Slava+Andreyev%2E+2010%2E+Using+entity%2Dbased+features+to+model+coherence+in+student+essays%2E+In+Proceedings+of+Human+Language+Technologies+2010%3A+The+Conference+of+the+Nor)
4. Donna K. Byron and Amanda Stent. 1998. A preliminary model of centering in dialog. In Proceedings of the 17th International Conference on Computational Linguistics and 36th Annual Meeting of the Association for Computational Linguistics, Montréal, Québec, Canada, 10–14 August 1998, pages 1475–1477. [↗](https://doi.org/10.21236/ada362244)
5. Alessandra Cervone, Evgeny Stepanov, and Giuseppe Riccardi. 2018. Coherence models for dialogue. In Proceedings of the 19th Annual Conference of the International Speech Communication Association, Hyderabad, 2–6 September 2018, pages 1011–1015. [find](https://scholar.google.com/scholar?q=Alessandra+Cervone%2C+Evgeny+Stepanov%2C+and+Giuseppe+Riccardi%2E+2018%2E+Coherence+models+for+dialogue%2E+In+Proceedings+of+the+19th+Annual+Conference+of+the+International+Speech+Communication+Association%2C+Hyd)
6. Emily Dinan, Varvara Logacheva, Valentin Malykh, Alexander H. Miller, Kurt Shuster, Jack Urbanek, Douwe Kiela, Arthur Szlam, Iulian Serban, Ryan Lowe, Shrimai Prabhumoye, Alan W. Black, Alexander I. Rudnicky, Jason Williams, Joelle Pineau, Mikhail Burtsev, and Jason Weston. 2019. [The second conversational intelligence challenge (convai2)](https://md.research.elamur.ai/free/en/papers/dinan-2019-second-conversational-intelligence-challenge-convai2/). CoRR, abs/1902.00098. [↗](https://arxiv.org/abs/1902.00098)
7. Nouha Dziri, Ehsan Kamalloo, Kory Mathewson, and Osmar Zaiane. 2019. Evaluating coherence in dialogue systems using entailment. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, Minnesota., 2–7 June 2019, pages 3806–3812. [find](https://scholar.google.com/scholar?q=Nouha+Dziri%2C+Ehsan+Kamalloo%2C+Kory+Mathewson%2C+and+Osmar+Zaiane%2E+2019%2E+Evaluating+coherence+in+dialogue+systems+using+entailment%2E+In+Proceedings+of+the+2019+Conference+of+the+North+American+Chapter+of+t)
8. Youmna Farag and Helen Yannakoudakis. 2019. Multi-task learning for coherence modeling. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Florence, Italy, 28 July – 2 August, 2019, pages 629–639. [find](https://scholar.google.com/scholar?q=Youmna+Farag+and+Helen+Yannakoudakis%2E+2019%2E+Multi%2Dtask+learning+for+coherence+modeling%2E+In+Proceedings+of+the+57th+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%28Volume+1%3A+Long+Paper)
9. Ziming Li, Julia Kiseleva, and Maarten de Rijke. 2019. Dialogue generation: From imitation learning to inverse reinforcement learning. In Proceedings of the 33rd Conference on the Advancement of Artificial Intelligence, Honolulu, Hawaii, 21 January –1 February 2019, pages 6722–6729. [find](https://scholar.google.com/scholar?q=Ziming+Li%2C+Julia+Kiseleva%2C+and+Maarten+de+Rijke%2E+2019%2E+Dialogue+generation%3A+From+imitation+learning+to+inverse+reinforcement+learning%2E+In+Proceedings+of+the+33rd+Conference+on+the+Advancement+of+Artif)
10. Sudeep Gandhe and David Traum. 2008. Evaluation understudy for dialogue coherence models. In Proceedings of the 9th SIGdial Workshop on Discourse and Dialogue, Columbus, Ohio, 19–20 June 2008, pages 172–181. [find](https://scholar.google.com/scholar?q=Sudeep+Gandhe+and+David+Traum%2E+2008%2E+Evaluation+understudy+for+dialogue+coherence+models%2E+In+Proceedings+of+the+9th+SIGdial+Workshop+on+Discourse+and+Dialogue%2C+Columbus%2C+Ohio%2C+19%E2%80%9320+June+2008%2C+pages+1)
11. Chia-Wei Liu, Ryan Lowe, Iulian Serban, Mike Nose-worthy, Laurent Charlin, and Joelle Pineau. 2016. [How NOT to evaluate your dialogue system: An empirical study of unsupervised evaluation metrics for dialogue response generation](https://md.research.elamur.ai/free/en/papers/liu-2016-how-not-evaluate-your-dialogue/). In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, Austin, Texas, 1–5 November 2016, pages 2122–2132.
12. Sudeep Gandhe and David Traum. 2016. A semi-automated evaluation metric for dialogue model coherence. In 7th International Workshop on Spoken Dialogue Systems, Saariselk”a, Finland, 13–16 January 2016, pages 141–150. [↗](https://doi.org/10.1007/978-3-319-21834-2_19)
13. Yang Gao, Christian M. Meyer, Mohsen Mesgar, and Iryna Gurevych. 2019. Reward learning for efficient reinforcement learning in extractive document summarisation. In Proceedings of the 28th International Joint Conference on Artificial Intelligence, Macao, China, 10–16 August 2019, pages 2350–2356. [find](https://scholar.google.com/scholar?q=Yang+Gao%2C+Christian+M%2E+Meyer%2C+Mohsen+Mesgar%2C+and+Iryna+Gurevych%2E+2019%2E+Reward+learning+for+efficient+reinforcement+learning+in+extractive+document+summarisation%2E+In+Proceedings+of+the+28th+Internation)
14. Marjan Ghazvininejad, Chris Brockett, Ming Wei Chang, Bill Dolan, Jianfeng Gao, Wen Tau Yih, and Michel Galley. 2018. A knowledge-grounded neural conversation model. In Proceedings of the 32ed Conference on the Advancement of Artificial Intelligence, New Orleans, Louisiana, 2–7 February 2018, pages 5110–5117. [↗](https://doi.org/10.1609/aaai.v32i1.11977)
15. Barbara J. Grosz and Candace L. Sidner. 1986. [Attention, intentions, and the structure of discourse](https://md.research.elamur.ai/free/en/papers/grosz-1986-attention-intentions-structure-discourse/). Computational Linguistics, 12(3):175–204. [↗](https://doi.org/10.5555/12457.12458)
16. Camille Guinaudeau and Michael Strube. 2013. Graph-based local coherence modeling. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Sofia, Bulgaria, 4–9 August 2013, pages 93–103. [find](https://scholar.google.com/scholar?q=Camille+Guinaudeau+and+Michael+Strube%2E+2013%2E+Graph%2Dbased+local+coherence+modeling%2E+In+Proceedings+of+the+51st+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%28Volume+1%3A+Long+Papers%29%2C+S)
17. M. A. K. Halliday and Ruqaiya Hasan. 1976. Cohesion in English. London, U.K.: Longman. [find](https://scholar.google.com/scholar?q=M%2E+A%2E+K%2E+Halliday+and+Ruqaiya+Hasan%2E+1976%2E+Cohesion+in+English%2E+London%2C+U%2EK%2E%3A+Longman%2E)
18. Daniel Jurafsky and Elizabeth Shriberg. 1997. Switchboard SWBD-DAMSL shallow-discourse-function annotation coders manual, draft 13. Technical Report 97-02, University of Colorado at Boulder. [find](https://scholar.google.com/scholar?q=Daniel+Jurafsky+and+Elizabeth+Shriberg%2E+1997%2E+Switchboard+SWBD%2DDAMSL+shallow%2Ddiscourse%2Dfunction+annotation+coders+manual%2C+draft+13%2E+Technical+Report+97%2D02%2C+University+of+Colorado+at+Boulder%2E)
19. Alex Kendall, Yarin Gal, and Roberto Cipolla. 2018. Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In Proceedings of the International Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, 18–22 June 2018, pages 7482–7491. [find](https://scholar.google.com/scholar?q=Alex+Kendall%2C+Yarin+Gal%2C+and+Roberto+Cipolla%2E+2018%2E+Multi%2Dtask+learning+using+uncertainty+to+weigh+losses+for+scene+geometry+and+semantics%2E+In+Proceedings+of+the+International+Conference+on+Computer+V)
20. Yanran Li, Hui Su, Xiaoyu Shen, Wenjie Li, Ziqiang Cao, and Shuzi Niu. 2017. [Dailydialog: A manually labelled multi-turn dialogue dataset](https://md.research.elamur.ai/free/en/papers/li-2017-dailydialog-manually-labelled-multi-turn/). In Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 1: Long Papers), Taipei, Taiwan, 27 November – 1 December, 2017, pages 986—-995. [↗](https://doi.org/10.48550/arxiv.1710.03957)
21. Mohsen Mesgar and Michael Strube. 2014. Normalized entity graph for computing local coherence. In Proceedings of TextGraphs-9: Graph-based Methods for Natural Language Processing, Workshop at EMNLP 2014, Doha, Qatar, 29 October 2014, pages 1–5. [find](https://scholar.google.com/scholar?q=Mohsen+Mesgar+and+Michael+Strube%2E+2014%2E+Normalized+entity+graph+for+computing+local+coherence%2E+In+Proceedings+of+TextGraphs%2D9%3A+Graph%2Dbased+Methods+for+Natural+Language+Processing%2C+Workshop+at+EMNLP+20)
22. Mohsen Mesgar and Michael Strube. 2018. A neural local coherence model for text quality assessment. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium, 31 October – 4 November 2018, pages 4328–4339. [find](https://scholar.google.com/scholar?q=Mohsen+Mesgar+and+Michael+Strube%2E+2018%2E+A+neural+local+coherence+model+for+text+quality+assessment%2E+In+Proceedings+of+the+2018+Conference+on+Empirical+Methods+in+Natural+Language+Processing%2C+Brussels%2C)
23. Jeffrey Pennington, Richard Socher, and Christopher Manning. 2014. [Glove: Global vectors for word representation](https://md.research.elamur.ai/free/en/papers/pennington-2014-glove-global-vectors-word-representation/). In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, Doha, Qatar, 25–29 October 2014, pages 1532– 1543.
24. C. Raymond Perrault and James F. Allen. 1978. Speech acts as a basis for understanding dialogue coherence. In Theoretical Issues in Natural Language Processing-2. [find](https://scholar.google.com/scholar?q=C%2E+Raymond+Perrault+and+James+F%2E+Allen%2E+1978%2E+Speech+acts+as+a+basis+for+understanding+dialogue+coherence%2E+In+Theoretical+Issues+in+Natural+Language+Processing%2D2%2E)
25. Amruta Purandare and Diane J. Litman. 2008. Analyzing dialog coherence using transition patterns in lexical and semantic features. In Proceedings of the 21st International Florida Artificial Intelligence Research Society Conference, Coconut Grove, Florida, 15–17 May 2008, pages 195–200. [find](https://scholar.google.com/scholar?q=Amruta+Purandare+and+Diane+J%2E+Litman%2E+2008%2E+Analyzing+dialog+coherence+using+transition+patterns+in+lexical+and+semantic+features%2E+In+Proceedings+of+the+21st+International+Florida+Artificial+Intellige)
26. Vipul Raheja and Joel Tetreault. 2019. [Dialogue act classification with context-aware self-attention](https://md.research.elamur.ai/free/en/papers/raheja-2019-dialogue-act-classification-context-aware/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, Minnesota., 2–7 June 2019, pages 3727–3733. [↗](https://doi.org/10.48550/arxiv.1904.02594)
27. Nils Reimers and Iryna Gurevych. 2018. Why comparing single performance scores does not allow to draw conclusions about machine learning approaches. CoRR, abs/1803.09578. [↗](https://arxiv.org/abs/1803.09578)
28. Gerard Salton. 1971. The SMART Retrieval System – Experiments in Automatic Document Processing. Englewood Cliffs, N.J.: Prentice Hall. [find](https://scholar.google.com/scholar?q=Gerard+Salton%2E+1971%2E+The+SMART+Retrieval+System+%E2%80%93+Experiments+in+Automatic+Document+Processing%2E+Englewood+Cliffs%2C+N%2EJ%2E%3A+Prentice+Hall%2E)
29. John Searle. 1969. [Speech Acts](https://md.research.elamur.ai/free/en/papers/searle-1969-speech-acts/). Cambridge University Press, Cambridge, U.K.
30. Iulian V Serban, Alessandro Sordoni, Yoshua Bengio, Aaron Courville, and Joelle Pineau. 2016. [Building end-to-end dialogue systems using generative hierarchical neural network models](https://md.research.elamur.ai/free/en/papers/serban-2015-building-end-end-dialogue-systems/). In Proceedings of the 30th Conference on the Advancement of Artificial Intelligence, Phoenix, Arizona, 12–17 February 2016, pages 3776–3783. [↗](https://doi.org/10.1609/aaai.v30i1.9883)
31. Dat Tien Nguyen and Shafiq Joty. 2017. A neural local coherence model. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Vancouver, Canada, 30 July – 4 August 2017, pages 1320–1330. [find](https://scholar.google.com/scholar?q=Dat+Tien+Nguyen+and+Shafiq+Joty%2E+2017%2E+A+neural+local+coherence+model%2E+In+Proceedings+of+the+55th+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%28Volume+1%3A+Long+Papers%29%2C+Vancouver%2C+Ca)
32. Svitlana Vakulenko, Maarten de Rijke, Michael Cochez, Vadim Savenkov, and Axel Polleres. 2018. Measuring semantic coherence of a conversation. In Proceedings of the 17th International Semantic Web Conference, Monterey, Ca., 8-12 October 2018, pages 634–651. [↗](https://doi.org/10.48550/arxiv.1806.06411)
33. Xinnuo Xu, Ondřej Dušek, Ioannis Konstas, and Verena Rieser. 2018. Better conversations by modeling, filtering, and optimizing for coherence and diversity. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium, 31 October – 4 November 2018, pages 3981–3991. [find](https://scholar.google.com/scholar?q=Xinnuo+Xu%2C+Ondr%CC%8Cej+Dus%CC%8Cek%2C+Ioannis+Konstas%2C+and+Verena+Rieser%2E+2018%2E+Better+conversations+by+modeling%2C+filtering%2C+and+optimizing+for+coherence+and+diversity%2E+In+Proceedings+of+the+2018+Conference+on+E)
34. Hainan Zhang, Yanyan Lan, Jiafeng Guo, Jun Xu, and Xueqi Cheng. 2018. Reinforcing coherence for sequence to sequence model in dialogue generation. In Proceedings of the 27th International Joint Conference on Artificial Intelligence, Stockholm, Sweden, 13–19 July 2018, pages 4567–4573. [find](https://scholar.google.com/scholar?q=Hainan+Zhang%2C+Yanyan+Lan%2C+Jiafeng+Guo%2C+Jun+Xu%2C+and+Xueqi+Cheng%2E+2018%2E+Reinforcing+coherence+for+sequence+to+sequence+model+in+dialogue+generation%2E+In+Proceedings+of+the+27th+International+Joint+Confer)

*Updated 24 October 2026*