---
title: "PoETIC: A Re-framing of Context Dependent Emotion Detection"
description: "They release PoETIC, a human-annotated benchmark built from filtered GoEmotions test sentences across six Ekman… 14 references."
authors: ["Nirmal Surange", "Manish Shrivastava"]
year: 2026
venue: "Proceedings of Computational Affective Science (CAS) @ LREC 2026"
source: "https://aclanthology.org/2026.cas-1.17/"
doi: "10.63317/4f24agmr53nm"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/surange-2026-poetic-re-framing-context-dependent/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/surange-2026-poetic-re-framing-context-dependent/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/45/) / PoETIC: A Re-framing of Context Dependent Emotion Detection

# PoETIC: A Re-framing of Context Dependent Emotion Detection

**Authors:** Nirmal Surange and [Manish Shrivastava](https://md.research.elamur.ai/free/en/authors/manish-shrivastava/) · **Year:** 2026 · **Venue:** [Proceedings of Computational Affective Science (CAS) @ LREC 2026](https://md.research.elamur.ai/free/en/venues/cas/) ([Computational Affective Science (CAS) @ LREC 2026](https://md.research.elamur.ai/free/en/venues/cas/2026/)) · **Publisher:** [European Language Resources Association (ELRA)](https://md.research.elamur.ai/free/en/publishers/european-language-resources-association-elra/) · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.cas-1.17/) · [PDF](https://aclanthology.org/2026.cas-1.17.pdf) · [DOI 10.63317/4f24agmr53nm](https://doi.org/10.63317/4f24agmr53nm) · [OpenAlex](https://openalex.org/W7166556335)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Most text-based emotion datasets label an utterance with emotions as if the feeling were fully contained in the words. The authors argue this is misleading. Without prosody, faces or emojis, written emotion depends heavily on the situation. Annotators labelling isolated sentences are therefore implicitly imagining a context. The paper recasts the task at the level of utterance–emotion pairs. Each pair is assigned one of three classes:
- **Emotion-Obvious**: the text conveys the emotion on its own.
- **Emotion-Plausible**: it could convey the emotion in some imagined situation.
- **Emotion-Implausible**: no context would make it fit.

They release PoETIC, a human-annotated benchmark built from filtered GoEmotions test sentences across six Ekman emotions plus Neutral. When asked to imagine contexts, annotators accept far more emotions per sentence than the original gold labels contain. The original GoEmotions tags also do not line up well with these plausibility judgments. Zero- and few-shot baselines with large and small language models are provided. The resource matters because it exposes how ambiguous single-label emotion annotation is. It also offers a way to test whether models reason about implicit context.

## Key points
- PoETIC contains about 7k utterance–emotion pairs: 1,000 sentences × 7 labels (Fear, Anger, Sadness, Joy, Disgust, Surprise, Neutral). It is released on GitHub (nirmalsurange/PoETIC).
- The sentences come from the GoEmotions test split. Filtering reduced its 5,427 sentences to 4,770 clean ones, from which 1,000 were selected. The removals were:
  - 268 sentences containing emojis
  - 267 with foul language
  - 106 short or noisy items
  - 16 with hashtags or URLs
- Annotators judged an average of 4.55 emotions per utterance as possible, versus 1.09 gold emotions per utterance in the GoEmotions test set.
- The three labels are abbreviated EO, EP and EI. For Neutral, the authors add a constraint: no sentence may receive the Emotion-Obvious label for Neutral and for some other emotion at the same time.
- Baselines cover three LLMs (GPT-5-mini, Gemini 2.5 Flash, Llama3-70B via Groq) and two small language models, each tested in zero-shot and few-shot settings.

*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

- [International Institute of Information Technology, Hyderabad](https://md.research.elamur.ai/free/en/universities/international-institute-of-information-technology-hyderabad/) — University, Hyderabad, India · 42 works · 2,317 citations

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [💭 Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/) · 1,649 works · [💭 Experimental and Cognitive Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/experimental-and-cognitive-psychology/) · 930 works · [💭 Social Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/social-psychology/) · 330 works

🧭 Topics (1): [😊 Sentiment & emotion](https://md.research.elamur.ai/free/en/topics/sentiment-emotion/) · 1,035 works

## Cite

Cite as: Nirmal Surange and Manish Shrivastava. 2026. PoETIC: A Re-framing of Context Dependent Emotion Detection. Proceedings of Computational Affective Science (CAS) @ LREC 2026.

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

```bibtex
@inproceedings{surange2026poetic,
  title = {{PoETIC: A Re-framing of Context Dependent Emotion Detection}},
  author = {Nirmal Surange and Manish Shrivastava},
  booktitle = {Proceedings of Computational Affective Science (CAS) @ LREC 2026},
  year = {2026},
  doi = {10.63317/4f24agmr53nm},
  url = {https://doi.org/10.63317/4f24agmr53nm}
}
```

- [← previous PodBench: A Comprehensive Benchmark for Instruction-Aware Audio-Oriented Podcast Script Generation](https://md.research.elamur.ai/free/en/papers/xu-2026-podbench-comprehensive-benchmark-instruction-aware/)
- [next → POINTERS at UZH Shared Task 2026: Reasoning Probes for Argumentation Mining in UN Resolutions](https://md.research.elamur.ai/free/en/papers/sen-2026-pointers-uzh-shared-task-2026/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/45/)

## References

14 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Anuj Attri, Arnav Attri, Suman Banerjee, Amey Patil, Muthusamy Chelliah, Nikesh Garera, and Pushpak Bhattacharyya. 2025. LLMs as architects and critics for multi-source opinion summarization. In Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Confere… [find](https://scholar.google.com/scholar?q=Anuj+Attri%2C+Arnav+Attri%2C+Suman+Banerjee%2C+Amey+Patil%2C+Muthusamy+Chelliah%2C+Nikesh+Garera%2C+and+Pushpak+Bhattacharyya%2E+2025%2E+LLMs+as+architects+and+critics+for+multi%2Dsource+opinion+summarization%2E+In+Proce)
2. Sven Büchel and Udo Hahn. 2017. Emobank: Studying the impact of annotation perspective and representation format on dimensional emotion analysis. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pages 578–585. [find](https://scholar.google.com/scholar?q=Sven+B%C3%BCchel+and+Udo+Hahn%2E+2017%2E+Emobank%3A+Studying+the+impact+of+annotation+perspective+and+representation+format+on+dimensional+emotion+analysis%2E+In+Proceedings+of+the+15th+Conference+of+the+European+)
3. Andiswa Bukula, Guanqun Cao, Rodrigo Tufiño, Rendi Chevi, Chiamaka Ijeoma Chukwuneke, Alexandra Ciobotaru, Daryna Dementieva, Murja Sani Gadanya, Robert Geislinger, Bela Gipp, Oumaima Hourrane, Oana Ignat, Falalu Ibrahim Lawan, Rooweither Mabuya, Rahmad Mahendra, Vukosi Marivate, Alexander Panchenko… [find](https://scholar.google.com/scholar?q=Andiswa+Bukula%2C+Guanqun+Cao%2C+Rodrigo+Tufi%C3%B1o%2C+Rendi+Chevi%2C+Chiamaka+Ijeoma+Chukwuneke%2C+Alexandra+Ciobotaru%2C+Daryna+Dementieva%2C+Murja+Sani+Gadanya%2C+Robert+Geislinger%2C+Bela+Gipp%2C+Oumaima+Hourrane%2C+Oana+I)
4. Carlos Busso, Murtaza Bulut, Chi-Chun Lee, Abe Kazemzadeh, Emily Mower, Samuel Kim, Jeannette N Chang, Sungbok Lee, and Shrikanth S Narayanan. 2008. Iemocap: Interactive emotional dyadic motion capture database. Language resources and evaluation, 42(4):335–359. [find](https://scholar.google.com/scholar?q=Carlos+Busso%2C+Murtaza+Bulut%2C+Chi%2DChun+Lee%2C+Abe+Kazemzadeh%2C+Emily+Mower%2C+Samuel+Kim%2C+Jeannette+N+Chang%2C+Sungbok+Lee%2C+and+Shrikanth+S+Narayanan%2E+2008%2E+Iemocap%3A+Interactive+emotional+dyadic+motion+captur)
5. Ankush Chatterjee, Kedhar Nath Narahari, Meghana Joshi, and Puneet Agrawal. 2019. [Semeval-2019 task 3: Emocontext contextual emotion detection in text](https://md.research.elamur.ai/free/en/papers/chatterjee-2019-semeval-2019-task-3-emocontext/). In Proceedings of the 13th international workshop on semantic evaluation, pages 39–48.
6. Dorottya Demszky, Dana Movshovitz-Attias, Jeongwoo Ko, Alan Cowen, Gaurav Nemade, and Sujith Ravi. 2020. GoEmotions: A dataset of fine-grained emotions. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4040–4054, Online. Association for Computational… [find](https://scholar.google.com/scholar?q=Dorottya+Demszky%2C+Dana+Movshovitz%2DAttias%2C+Jeongwoo+Ko%2C+Alan+Cowen%2C+Gaurav+Nemade%2C+and+Sujith+Ravi%2E+2020%2E+GoEmotions%3A+A+dataset+of+fine%2Dgrained+emotions%2E+In+Proceedings+of+the+58th+Annual+Meeting+of+th)
7. Paul Ekman. 1992. Are there basic emotions? Psychological Review, 99(3):550–553. [find](https://scholar.google.com/scholar?q=Paul+Ekman%2E+1992%2E+Are+there+basic+emotions%3F+Psychological+Review%2C+99%283%29%3A550%E2%80%93553%2E)
8. Roman Klinger, Orphée De Clercq, Saif Mohammad, and Alexandra Balahur. 2018. IEST: WASSA-2018 implicit emotions shared task. In Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis, pages 31–42, Brussels, Belgium. Association for Computatio… [find](https://scholar.google.com/scholar?q=Roman+Klinger%2C+Orph%C3%A9e+De+Clercq%2C+Saif+Mohammad%2C+and+Alexandra+Balahur%2E+2018%2E+IEST%3A+WASSA%2D2018+implicit+emotions+shared+task%2E+In+Proceedings+of+the+9th+Workshop+on+Computational+Approaches+to+Subjectiv)
9. Florian Lecourt, Madalina Croitoru, and Konstantin Todorov. 2025. ’only chatgpt gets me’: An empirical analysis of gpt versus other large language models for emotion detection in text. In Companion Proceedings of the ACM on Web Conference 2025, pages 2603–2611. [find](https://scholar.google.com/scholar?q=Florian+Lecourt%2C+Madalina+Croitoru%2C+and+Konstantin+Todorov%2E+2025%2E+%E2%80%99only+chatgpt+gets+me%E2%80%99%3A+An+empirical+analysis+of+gpt+versus+other+large+language+models+for+emotion+detection+in+text%2E+In+Companion+Pr)
10. Gary McKeown, Michel Valstar, Roddy Cowie, Maja Pantic, and Marc Schroder. 2011. The semaine database: Annotated multimodal records of emotionally colored conversations between a person and a limited agent. IEEE transactions on affective computing, 3(1):5–17. [find](https://scholar.google.com/scholar?q=Gary+McKeown%2C+Michel+Valstar%2C+Roddy+Cowie%2C+Maja+Pantic%2C+and+Marc+Schroder%2E+2011%2E+The+semaine+database%3A+Annotated+multimodal+records+of+emotionally+colored+conversations+between+a+person+and+a+limited+)
11. Shamsuddeen Hassan Muhammad, Nedjma Ousidhoum, Idris Abdulmumin, Jan Philip Wahle, Terry Ruas, Meriem Beloucif, Christine de Kock, Nirmal Surange, Daniela Teodorescu, Ibrahim Said Ahmad, David Ifeoluwa Adelani, Alham Fikri Aji, Felermino D. M. A. Ali, Ilseyar Alimova, Vladimir Araujo, Nikolay Babako… — [MELD: A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations](https://md.research.elamur.ai/free/en/papers/poria-2018-meld-multimodal-multi-party-dataset/)
12. Klaus R Scherer and Harald G Wallbott. 1994. Evidence for universality and cultural variation of differential emotion response patterning. Journal of personality and social psychology, 66(2):310. [find](https://scholar.google.com/scholar?q=Klaus+R+Scherer+and+Harald+G+Wallbott%2E+1994%2E+Evidence+for+universality+and+cultural+variation+of+differential+emotion+response+patterning%2E+Journal+of+personality+and+social+psychology%2C+66%282%29%3A310%2E)
13. Smriti Singh, Cornelia Caragea, and Junyi Jessy Li. 2024. Language models (mostly) do not consider emotion triggers when predicting emotion. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Sh… [find](https://scholar.google.com/scholar?q=Smriti+Singh%2C+Cornelia+Caragea%2C+and+Junyi+Jessy+Li%2E+2024%2E+Language+models+%28mostly%29+do+not+consider+emotion+triggers+when+predicting+emotion%2E+In+Proceedings+of+the+2024+Conference+of+the+North+American)
14. Daniel Yang, Aditya Kommineni, Mohammad Alshehri, Nilamadhab Mohanty, Vedant Modi, Jonathan Gratch, and Shrikanth Narayanan. 2023. Context unlocks emotions: Text-based emotion classification dataset auditing with large language models. In 2023 11th International Conference on Affective Computing and… [find](https://scholar.google.com/scholar?q=Daniel+Yang%2C+Aditya+Kommineni%2C+Mohammad+Alshehri%2C+Nilamadhab+Mohanty%2C+Vedant+Modi%2C+Jonathan+Gratch%2C+and+Shrikanth+Narayanan%2E+2023%2E+Context+unlocks+emotions%3A+Text%2Dbased+emotion+classification+dataset+a)

*Updated 2 November 2026*