# L3Cube-MahaEmotions: A Marathi Emotion Recognition Dataset with Synthetic Annotations using CoTR prompting and Large Language Models

arXiv: https://arxiv.org/abs/2506.00863 · HTML: https://mlsanity.com/paper/2506.00863/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 25 references that are in the map.
- First posted on arXiv 1 Jun 2025.

## Builds on (most cited first, 12 of 12)

- [Don't Trust ChatGPT when Your Question is not in English: A Study of Multilingual Abilities and Types of LLMs](https://mlsanity.com/paper/2305.16339/) (2023)
- [Low-resource Languages: A Review of Past Work and Future Challenges](https://mlsanity.com/paper/2006.07264/) (2020)
- [L3Cube-MahaCorpus and MahaBERT: Marathi Monolingual Corpus, Marathi BERT Language Models, and Resources](https://mlsanity.com/paper/2202.01159/) (2022)
- [L3CubeMahaSent: A Marathi Tweet-based Sentiment Analysis Dataset](https://mlsanity.com/paper/2103.11408/) (2021)
- [L3Cube-MahaNLP: Marathi Natural Language Processing Datasets, Models, and Library](https://mlsanity.com/paper/2205.14728/) (2022)
- [L3Cube-MahaHate: A Tweet-based Marathi Hate Speech Detection Dataset and BERT models](https://mlsanity.com/paper/2203.13778/) (2022)
- [Establishing Baselines for Text Classification in Low-Resource Languages](https://mlsanity.com/paper/2005.02068/) (2020)
- [On Limitations of LLM as Annotator for Low Resource Languages](https://mlsanity.com/paper/2411.17637/) (2024)
- [Towards Emotion Recognition in Hindi-English Code-Mixed Data: A Transformer Based Approach](https://mlsanity.com/paper/2102.09943/) (2021)
- [L3Cube-MahaSent-MD: A Multi-domain Marathi Sentiment Analysis Dataset and Transformer Models](https://mlsanity.com/paper/2306.13888/) (2023)
- [EmoInHindi: A Multi-label Emotion and Intensity Annotated Dataset in Hindi for Emotion Recognition in Dialogues](https://mlsanity.com/paper/2205.13908/) (2022)
- [Emotion Classification in Low and Moderate Resource Languages](https://mlsanity.com/paper/2402.18424/) (2024)

## Built on by (most cited first, 2 of 2)

- [L3Cube-MahaPOS: A Marathi Part-of-Speech Tagging Dataset and BERT Models](https://mlsanity.com/paper/2606.24825/) (2026)
- [Enhancing Zero-Shot Domain Transfer in Low-Resource Sentiment Analysis: A Case Study in Marathi](https://mlsanity.com/paper/2610.15757/) (2026)

## Abstract (arXiv)

Emotion recognition in low-resource languages like Marathi remains challenging due to limited annotated data. We present L3Cube-MahaEmotions, a high-quality Marathi emotion recognition dataset with 11 fine-grained emotion labels. The training data is synthetically annotated using large language models (LLMs), while the validation and test sets are manually labeled to serve as a reliable gold-standard benchmark. Building on the MahaSent dataset, we apply the Chain-of-Translation (CoTR) prompting technique, where Marathi sentences are translated into English and emotion labeled via a single prompt. GPT-4 and Llama3-405B were evaluated, with GPT-4 selected for training data annotation due to superior label quality. We evaluate model performance using standard metrics and explore label aggregation strategies (e.g., Union, Intersection). While GPT-4 predictions outperform fine-tuned BERT models, BERT-based models trained on synthetic labels fail to surpass GPT-4. This highlights both the importance of high-quality human-labeled data and the inherent complexity of emotion recognition. An important finding of this work is that generic LLMs like GPT-4 and Llama3-405B generalize better than fine-tuned BERT for complex low-resource emotion recognition tasks. The dataset and model are shared publicly at https://github.com/l3cube-pune/MarathiNLP