CFP: BHASHA Workshop co-located with IJCNLP-AACL 2025

[Saptarshi Ghosh]

[apologies for cross-posting]

BHASHA Workshop: Benchmarks, Harmonization, Annotation, and Standardization for Human-Centric AI in Indian Languages
to be held with IJCNLP-AACL 2025 at IIT Bombay, Mumbai, India on 23rd December 2025
https://bhasha-workshop.github.io/


The BHASHA workshop aims to bring together researchers and practitioners working on datasets, benchmarks, resources, annotation, evaluation, and evaluation standards for Indian languages. We warmly encourage submissions from researchers, engineers, and language-technology practitioners who are actively working on resources, evaluation, and human-centric approaches for Indian languages.

Please feel free to forward this CFP to interested colleagues and acquaintances who might be interested.

== Scope for paper submissions ==

We solicit original, unpublished work on topics including (but not limited to):
* Creation and evaluation of datasets/benchmarks for Indian languages
* Annotation schemes, guidelines, and tooling for consistent labeling
* Harmonization and standardization of formats and metrics across Indian languages
* Resource collection, curation, and best practices for low-resource Indian languages
* Language-specific modeling challenges (morphology, free word order, script issues)
* Human-centric evaluation, fairness, accessibility, and usability studies
* Cross-lingual transfer, multilingual models, and evaluation for Indian languages
* Ethical, legal, and community-driven approaches to dataset creation and release
* Methods for domain-specific tasks for India in Indian languages (e.g., in Healthcare, Law, Education)

We also invite submissions describing system reports, negative results, and position papers that advance the workshop goals.


== Shared Tasks ==

BHASHA includes two shared tasks on Grammatical Error Detection and Correction for Indian languages (IndicGEC 2025) and Word Group Identification in Indian Languages. Details, task subtasks, and evaluation metrics (Macro-F1, GLEU) are on the workshop shared-task tab (https://bhasha-workshop.github.io/sharedtask.html). Organizers encourage participants to join the shared task and submit system papers.


== Submission Instructions ==

Peer-reviewed submissions are accepted via ARR (ACL Rolling Review) or via direct submission.  Please follow the detailed instructions on the workshop website for the preferred format, anonymization policy, and templates. If you plan to submit a shared-task system paper, please follow the shared-task timeline and submission instructions available on the shared task page.

We also look for non-archival submissions (e.g., previously accepted papers or Arxiv papers) and demonstrations that will be presented in the workshop.


== Important dates (Anywhere on Earth) ==

* Paper submission deadline (for direct submissions): October 10, 2025
* ARR commitment deadline (if submitting via ARR): October 30, 2025
* Non-archival and demonstration submission deadline: October 10, 2025
* Notification of acceptance: November 5, 2025
* Camera-ready due: November 15, 2025

The timeline for the shared task (training/test data release, and system paper deadlines) is listed on the shared task page.


== Contact ==

For more details, please visit the workshop website at https://bhasha-workshop.github.io/.
For questions, contact the organizers at bhashaw...@gmail.com.

Best regards,
The BHASHA Organizing Committee
(IIT Kanpur, IIT Kharagpur, IISER Kolkata, IIT Bombay, BITS Pilani and collaborators)