---
title: "Proceedings of the 1st Workshop on — Bhattacharya 2025"
description: "These are the proceedings of the first BHASHA 2025 workshop."
authors: ["Arnab Bhattacharya", "Pawan Goyal", "Saptarshi Ghosh", "Kripabandhu Ghosh"]
year: 2025
source: "https://aclanthology.org/2025.bhasha-1.0/"
doi: "10.18653/v1/2025.bhasha-1.0"
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/papers/anon-2025-proceedings-1st-workshop-benchmarks-harmonization/"
---
[Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/anon-2025-proceedings-1st-workshop-benchmarks-harmonization/)

[home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [2025](https://md.research.elamur.ai/free/papers/2025/page/2/) / Proceedings of the 1st Workshop on Benchmarks, Harmonization, Annotation, and Standardization for Human-Centric AI in Indian Languages (BHASHA 2025)

# Proceedings of the 1st Workshop on Benchmarks, Harmonization, Annotation, and Standardization for Human-Centric AI in Indian Languages (BHASHA 2025)

**Authors:** [Arnab Bhattacharya](https://md.research.elamur.ai/free/authors/arnab-bhattacharya/), Pawan Goyal, Saptarshi Ghosh, [Kripabandhu Ghosh](https://md.research.elamur.ai/free/authors/kripabandhu-ghosh/) · **Year:** 2025 · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2025.bhasha-1.0/) · [PDF](https://aclanthology.org/2025.bhasha-1.0.pdf) · [DOI 10.18653/v1/2025.bhasha-1.0](https://doi.org/10.18653/v1/2025.bhasha-1.0) · [OpenAlex](https://openalex.org/W7137927808)

**Citations:** 0 citations (openalex-snapshot · 2026-10-14)

## Summary
These are the proceedings of the first BHASHA 2025 workshop. The workshop took place on 23 December 2025 at IIT Bombay as part of the IJCNLP-AACL 2025 conference. The organizers note that India has 22 official languages, yet NLP research for them lags behind. These languages are inflectional and agglutinative and have free word order, so tools built for other languages transfer poorly to them. The workshop therefore focuses on benchmarks, annotated corpora, resources and evaluation metrics for Indian languages. The proceedings include 11 peer-reviewed papers. They cover OCR and handwriting recognition for Devanagari, the semantic bank USR Bank 1.0, new evaluation sets for Hindi LLMs, hallucination and jailbreak-attack benchmarks, and accent restoration in Vedic Sanskrit. Two shared tasks are presented separately: grammatical error correction (IndicGEC) and word grouping (IndicWG). They come with an overview paper and participants' system descriptions. For Mova, the proceedings are useful as an example of how to build evaluation infrastructure for low-resource languages with complex morphology.

## Key points
- 26 papers were submitted; one was desk-rejected, and 11 were accepted (about 42%). Eight of them were presented orally. Eight of the accepted papers are from India, and one each from Japan, Canada and the USA. The program committee had 19 members.
- IndicGEC covered five languages: Hindi, Bengali, Telugu, Tamil and Malayalam. IndicWG was run for Hindi only. 14 and 2 teams, respectively, competed in the final stage. Of the 10 submitted system papers, 6 were accepted.
- New resources: USR Bank 1.0, inspired by the Indian grammatical tradition; AnciDev, manuscripts in ancient Devanagari; BHRAM-IL, hallucinations in several Indian languages; Mātṛkā, multilingual jailbreak-attack testing of open LLMs; a new set of benchmarks for Hindi LLMs.
- Participants' approaches: mT5, an asymmetric BPE configuration, a hybrid neurosymbolic method for Tamil and Malayalam, and instruction tuning with minimal edits.
- Two invited talks. Monojit Choudhury spoke about the "metacultural" competence of models. Kalika Bali spoke about evaluating AI with the participation of language communities, using the Samiksha and DOSA initiatives as examples.

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

Fields: [Linguistics & languages](https://md.research.elamur.ai/free/shelves/linguistics/)

## Cite

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

```bibtex
@misc{bhattacharya2025proceedings,
  title = {{Proceedings of the 1st Workshop on Benchmarks, Harmonization, Annotation, and Standardization for Human-Centric AI in Indian Languages (BHASHA 2025)}},
  author = {Bhattacharya, Arnab and Goyal, Pawan and Ghosh, Saptarshi and Ghosh, Kripabandhu},
  year = {2025},
  doi = {10.18653/v1/2025.bhasha-1.0},
  url = {https://doi.org/10.18653/v1/2025.bhasha-1.0}
}
```

- [← previous Proceedings of Bridging Neurons and Symbols for Natural Language Processing and Knowledge Graphs Reasoning @ COLING 2025](https://md.research.elamur.ai/free/papers/anon-2025-proceedings-bridging-neurons-symbols-natural/)
- [next → Proceedings of the Fourth Workshop on Bridging Human-Computer Interaction and Natural Language Processing (HCI+NLP)](https://md.research.elamur.ai/free/papers/anon-2025-proceedings-fourth-workshop-bridging-human/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/papers/2025/page/2/)