Elamur Research⌕
☰
home / papers / 2026 / Emotion Trajectory-aware Retrieval for Markov-driven Emotion Anticipation in LLM-based Emotional Support Conversation

Emotion Trajectory-aware Retrieval for Markov-driven Emotion Anticipation in LLM-based Emotional Support Conversation

Paper · 2026 · Hongyan Wu, Zhiliang Tian, Zhen Huang et al.

Summary

This paper addresses strategy selection in LLM-based emotional support conversation (ESC). Existing planners mostly react to the current dialogue state. Retrieval-based anticipation methods match only the user's present emotion and ignore how that emotion evolved over earlier turns. The authors propose a framework with three parts:

  • Dynamic emotion memory. Built from UniConv training dialogues, it stores lexicon-derived VAD emotion trajectories (from NRC_VAD), discrete emotion labels and dialogue metadata. Retrieval is hierarchical, combining semantic similarity with alignment of emotion trajectories.
  • Markov-driven anticipation. Emotion transitions are treated as a Markov chain. The transition matrix is updated using emotion changes estimated from retrieved trajectories, adjusted by a user personality vector.
  • Strategy planning. The anticipated emotion prompts an LLM to propose candidate strategies. A planner is then trained with RL-based active online learning, which samples difficult user personas more often.

Across two datasets and two backbone models, the method beats all baselines. It also sustains more positive user emotion in later dialogue turns. The work suggests that modelling emotional history, rather than isolated states, yields better lookahead support strategies.

Key points

  • Emotion trajectories come from sliding-window aggregation of word-level Valence-Arousal-Dominance scores from the NRC_VAD lexicon over UniConv dialogues. EXTES is used as an additional evaluation dataset.
  • Three retrieval orderings were compared: semantic-then-emotion, emotion-then-semantic and score fusion. Emotion-first (E-S) retrieval performed best overall.
  • Ablations remove the Markov anticipation, active learning, RL optimization or the personality vector. Each removal lowers performance.
  • In the persona analysis (BERT-encoded strategy sequences, reduced with t-SNE), the method had the lowest Intra-Persona distance (13.10) and the highest Inter-Persona distance (29.88). The comparison included ICL-AIF, PPDPP, TRIP, PRINCIPLES and DialogXpert.
  • Stated limitations: the method is text-only and does not guarantee relief of negative emotions. It is not meant to replace professional counseling.

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

🎓National University of Defense TechnologyChangsha · China20 works · 1,339 citations🎓Chongqing UniversityChongqing · China14 works · 318 citations

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksPsychology · 1,649 worksExperimental and Cognitive Psychology · 930 works

🧭 Topics 2Research subjects and languages the work is about.Sentiment & emotion · 1,035 worksInformation retrieval & search · 552 works

Cite

Cite as Hongyan Wu, Zhiliang Tian, Zhen Huang, Tengyue Hu, Linbo Qiao, Yifu Gao, Feng Liu and Dongsheng Li. 2026. Emotion Trajectory-aware Retrieval for Markov-driven Emotion Anticipation in LLM-based Emotional Support Conversation. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.2127

@inproceedings{wu2026emotion,
  title = {{Emotion Trajectory-aware Retrieval for Markov-driven Emotion Anticipation in LLM-based Emotional Support Conversation}},
  author = {Hongyan Wu, Zhiliang Tian, Zhen Huang, Tengyue Hu, Linbo Qiao, Yifu Gao, Feng Liu and Dongsheng Li},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.2127},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.2127}
}

References

51 references, 8 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. M. S. Bartlett. 1937. Properties of sufficiency and statistical tests. Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences, 160(901):268–282. find
  2. Yaru Cao, Zhuang Chen, Guanqun Bi, Yulin Feng, Min Chen, Fucheng Wan, Minlie Huang, and Hongzhi Yu. 2024. Enhancing emotional support conversation with cognitive chain-of-thought reasoning. In CCF International Conference on Natural Language Processing and Chinese Computing, pages 175–187. Springer. find
  3. Yirong Chen, Xiaofen Xing, Jingkai Lin, Huimin Zheng, Zhenyu Wang, Qi Liu, and Xiangmin Xu. 2023. Soulchat: Improving llms’ empathy, listening, and comfort abilities through fine-tuning with multi-turn empathy conversations. In Findings of the Association for Computational Linguistics: EMNLP 2023, p… find
  4. Zhuang Chen, Yaru Cao, Guanqun Bi, Jincenzi Wu, Jinfeng Zhou, Xiyao Xiao, Si Chen, Hongning Wang, and Minlie Huang. 2025. Socialsim: Towards socialized simulation of emotional support conversation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 1274–1282. find
  5. Jiale Cheng, Sahand Sabour, Hao Sun, Zhuang Chen, and Minlie Huang. 2023. Pal: Persona-augmented emotional support conversation generation. In Findings of the Association for Computational Linguistics: ACL 2023, pages 535–554. find
  6. Yi Cheng, Wenge Liu, Wenjie Li, Jiashuo Wang, Ruihui Zhao, Bang Liu, Xiaodan Liang, and Yefeng Zheng. 2022. Improving multi-turn emotional support dialogue generation with lookahead strategy planning. arXiv preprint arXiv:2210.04242. find
  7. Pietro Cipresso, Francesca Borghesi, and Alice Chirico. 2023. Affects affect affects: A markov chain. Frontiers in Psychology, 14:1162655. find
  8. Kate Daley, Ines Hungerbuehler, Kate Cavanagh, Heloísa Garcia Claro, Paul Alan Swinton, and Michael Kapps. 2020. Preliminary evaluation of the engagement and effectiveness of a mental health chatbot. Frontiers Digit. Health, 2:576361. find
  9. Yang Deng, Lizi Liao, Liang Chen, Hongru Wang, Wenqiang Lei, and Tat-Seng Chua. 2023. Prompting and evaluating large language models for proactive dialogues: Clarification, target-guided, and non-collaboration. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 10602–106…
  10. Yang Deng, Wenxuan ZHANG, Wai LAM, See-Kiong NG, and Tat-Seng CHUA. 2024. Plug-and-play policy planner for large language model powered dialogue agents. find
  11. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies,…
  12. Faiza Farhat. 2024. Chatgpt as a complementary mental health resource: a boon or a bane. Annals of Biomedical Engineering, 52(5):1111–1114. find
  13. Yao Fu, Hao Peng, Tushar Khot, and Mirella Lapata. 2023. Improving language model negotiation with self-play and in-context learning from ai feedback. arXiv preprint arXiv:2305.10142.
  14. Lewis R Goldberg. 1992. The development of markers for the big-five factor structure. Psychological assessment, 4(1):26. find
  15. Jiawang Hao and Fang Kong. 2025. Enhancing emotional support conversations: A framework for dynamic knowledge filtering and persona extraction. In Proceedings of the 31st International Conference on Computational Linguistics, pages 3193–3202. find
  16. Tao He, Lizi Liao, Yixin Cao, Yuanxing Liu, Ming Liu, Zerui Chen, and Bing Qin. 2024. Planning like human: A dual-process framework for dialogue planning. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4768–4791, Bangkok, Tha… find
  17. Tao He, Lizi Liao, Yixin Cao, Yuanxing Liu, Yiheng Sun, Zerui Chen, Ming Liu, and Bing Qin. 2025a. Simulation-free hierarchical latent policy planning for proactive dialogues. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 24032–24040. find
  18. Tao He, Lizi Liao, Ming Liu, and Bing Qin. 2025b. Simulating before planning: Constructing intrinsic user world model for user-tailored dialogue policy planning. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 645–655. find
  19. Le Huang, Hengzhi Lan, Zijun Sun, Chuan Shi, and Ting Bai. 2024. Emotional rag: Enhancing role-playing agents through emotional retrieval. In 2024 IEEE International Conference on Knowledge Graph (ICKG), pages 120–127. IEEE. find
  20. Mengzhao Jia, Qianglong Chen, Liqiang Jing, Dawei Fu, and Renyu Li. 2024. Knowledge-enhanced memory model for emotional support conversation. In ML4CMH@ AAAI. find
  21. Zhengda Jin, Bingbing Wang, Geng Tu, and Ruifeng Xu. 2025. Overview of the nlpcc 2025 shared task 8: Personalized emotional support conversation. In CCF International Conference on Natural Language Processing and Chinese Computing, pages 512–520. Springer. find
  22. Dongjin Kang, Sunghwan Mac Kim, Taeyoon Kwon, Seungjun Moon, Hyunsouk Cho, Youngjae Yu, Dongha Lee, and Jinyoung Yeo. 2024. Can large language models be good emotional supporter? mitigating preference bias on emotional support conversation. In Proceedings of the 62nd Annual Meeting of the Associatio… find
  23. Namyoung Kim, Kai Tzu-iunn Ong, Yeonjun Hwang, Minseok Kang, Iiseo Jihn, Gayoung Kim, Minju Kim, and Jinyoung Yeo. 2025. Principles: Synthetic strategy memory for proactive dialogue agents. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 21329–21368. find
  24. Yi-Feng Lin and Arbee LP Chen. 2025. Predicting the following support strategy during the emotional support dialogue by integrating the dialogue history and supporter features. Multimedia Tools and Applications, 84(24):28353–28373. find
  25. Siyang Liu, Chujie Zheng, Orianna Demasi, Sahand Sabour, Yu Li, Zhou Yu, Yong Jiang, and Minlie Huang. 2021. Towards emotional support dialog systems. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Lan…
  26. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692.
  27. Saif M Mohammad. 2025. Nrc vad lexicon v2: Norms for valence, arousal, and dominance for over 55k english terms. arXiv preprint arXiv:2503.23547.
  28. Robert W Norton. 1978. Foundation of a communicator style construct. Human communication research, 4(2):99–112. find
  29. Emily Ohman, Yuri Bizzoni, Pascale Feldkamp Moreira, and Kristoffer Nielbo. 2024. EmotionArcs: Emotion arcs for 9,000 literary texts. In Proceedings of the 8th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature (LaTeCH-CLfL 2024), pag… find
  30. Tazeek Bin Abdur Rakib, Ambuj Mehrish, Lay-Ki Soon, Wern Han Lim, and Soujanya Poria. 2025. Dialogxpert: Driving intelligent and emotion-aware conversations through online value-based reinforcement learning with llm priors. arXiv preprint arXiv:2505.17795. — [arXiv:2505.17795](https://arxiv.org/abs/2505.17795) find
  31. James A Russell. 1980. A circumplex model of affect. Journal of personality and social psychology, 39(6):1161. find
  32. Yixuan Tang, Yi Yang, and Ahmed Abbasi. 2025. Per-sonafuse: A personality activation-driven framework for enhancing human-llm interactions. arXiv preprint arXiv:2509.07370. — [arXiv:2509.07370](https://arxiv.org/abs/2509.07370) find
  33. Daniela Teodorescu and Saif Mohammad. 2023. Evaluating emotion arcs across languages: Bridging the global divide in sentiment analysis. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 4124–4137, Singapore. Association for Computational Linguistics.
  34. Zhiliang Tian, Wei Bi, Dongkyu Lee, Lanqing Xue, Yiping Song, Xiaojiang Liu, and Nevin L. Zhang. 2020. Response-anticipated memory for on-demand knowledge integration in response generation. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 650–659, On… find
  35. Quan Tu, Yanran Li, Jianwei Cui, Bin Wang, Ji-Rong Wen, and Rui Yan. 2022. Misc: A mixed strategy-aware model integrating comet for emotional support conversation. arXiv preprint arXiv:2203.13560.
  36. Chenwei Wan, Matthieu Labeau, and Chloé Clavel. 2025. Emodynamix: Emotional support dialogue strategy prediction by modelling mixed emotions and discourse dynamics. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human La… find
  37. Xiaoyu Wang, Yue Zhao, Qingqing Gu, Zhonglin Jiang, Yong Chen, and Luo Ji. 2025a. Convert language model into a value-based strategic planner. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track), pages 1444–1456. find
  38. Yunxiao Wang, Meng Liu, Wenqi Liu, Kaiyu Jiang, Bin Wen, Fan Yang, Tingting Gao, Guorui Zhou, and Liqiang Nie. 2025b. Compeer: Controllable empathetic reinforcement reasoning for emotional support conversation. arXiv preprint arXiv:2508.09521. — [arXiv:2508.09521](https://arxiv.org/abs/2508.09521) find
  39. Shenghan Wu, Yimo Zhu, Wynne Hsu, Mong-Li Lee, and Yang Deng. 2025. From personas to talks: Revisiting the impact of personas on llm-synthesized emotional support conversations. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 5439–5453. find
  40. Zhe Xu, Daoyuan Chen, Jiayi Kuang, Zihao Yi, Yaliang Li, and Ying Shen. 2024. Dynamic demonstration retrieval and cognitive understanding for emotional support conversation. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 774–… find
  41. Xiao Yu, Maximillian Chen, and Zhou Yu. 2023. Prompt-based monte-carlo tree search for goal-oriented dialogue policy planning. arXiv preprint arXiv:2305.13660. find
  42. Chao Zhang, Xin Shi, Xueqiao Zhang, Yifan Zhu, Yi Yang, and Yawei Luo. 2025a. Decoupledesc: Enhancing emotional support generation via strategy-response decoupled preference optimization. arXiv preprint arXiv:2505.16995. find
  43. Tong Zhang, Chen Huang, Yang Deng, Hongru Liang, Jia Liu, Zujie Wen, Wenqiang Lei, and Tat-Seng Chua. 2024. Strength lies in differences! improving strategy planning for non-collaborative dialogues via diversified user simulation. arXiv preprint arXiv:2403.06769. find
  44. Xinjie Zhang, Wenxuan Wang, and Qin Jin. 2025b. Intentionesc: An intention-centered framework for enhancing emotional support in dialogue systems. arXiv preprint arXiv:2506.05947. find
  45. Suifeng Zhao, Tong Zhou, Zhuoran Jin, Hongbang Yuan, Yubo Chen, Kang Liu, and Sujian Li. 2024. Awecita: Generating answer with appropriate and well-grained citations using llms. DATA INTELLI-GENCE, 6(4):1134–1157. find
  46. Weixiang Zhao, Xingyu Sui, Xinyang Han, Yang Deng, Yulin Hu, Jiahe Guo, Libo Qin, Qianyun Du, Shijin Wang, Yanyan Zhao, et al. 2025. Chain of strategy optimization makes large language models better emotional supporter. arXiv preprint arXiv:2503.05362. find
  47. Weixiang Zhao, Yanyan Zhao, Shilong Wang, and Bing Qin. 2023. Transesc: smoothing emotional support conversation via turn-level state transition. arXiv preprint arXiv:2305.03296. find
  48. Zhonghua Zheng, Lizi Liao, Yang Deng, and Liqiang Nie. 2023. Building emotional support chatbots in the era of llms. arXiv preprint arXiv:2308.11584. — [arXiv:2308.11584](https://arxiv.org/abs/2308.11584) find
  49. Nana Zhu, Caihai Zhu, Qingfu Zhu, and Yuanxing Liu. 2025. Style rewriter: A three-stage approach for stylistic generation of open-domain conversational responses. DATA INTELLIGENCE, 7(2):397–415. find
  50. Ting Yang, Li Chen, and Huimin Wang. 2025. Towards open-ended emotional support conversations in llms via reinforcement learning with future-oriented rewards. arXiv preprint arXiv:2508.12935. — [arXiv:2508.12935](https://arxiv.org/abs/2508.12935) find
  51. Jiong Yu, Sixing Wu, Jiahao Chen, and Wei Zhou. 2024. Llms as collaborator: Demands-guided collaborative retrieval-augmented generation for commonsense knowledge-grounded open-domain dialogue systems. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 13586– 13612. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wu-2026-emotion-trajectory-aware-retrieval-markov/