COMI-LINGUA: Expert Annotated Large-Scale Dataset for Multitask NLP in Hindi-English Code-Mixing
Sheth, Beniwal, Singh. Findings of EMNLP 2025.
PaperarXivMultilingual & Multimodal Intelligence
Code-mixed NLP
ANRF-supported benchmarks, annotation tools, and models for Hindi–English code-mixing (language identification, POS, NER, translation, and related tasks), February 2023 – February 2026.
- Agency
- ANRF
- Period / amount
- Feb 2023 – Feb 2026 · ₹47.674 lakh
Publications
Commentator: A Code-mixed Multilingual Text Annotation Framework
Sheth, Nisar, Prajapati, Beniwal, Singh. EMNLP 2024 System Demonstrations.
PaperGitHubDatasets
COMI-LINGUA
Expert-annotated Hindi–English code-mixing (181k+ instances) for LID, POS, NER, normalisation, and translation. Findings of EMNLP 2025.
PagePaperarXivGitHubHugging FaceModels
Task models trained on COMI-LINGUA. Weights on Hugging Face.
COMI-LINGUA-LID
Language identification (token classification).
Hugging FaceCOMI-LINGUA-POS
POS tagging.
Hugging FaceCOMI-LINGUA-NER
Named-entity recognition.
Hugging FaceCOMI-LINGUA-MLI
Matrix-language identification.
Hugging FaceCOMI-LINGUA-MT
Code-mixed translation.
Hugging FaceCOMI-LINGUA-TN
Text normalisation.
Hugging FaceCommentator
Code-mixed multilingual text annotation framework. EMNLP 2024 System Demonstrations.
Commentator: A Code-mixed Multilingual Text Annotation Framework
Sheth, Nisar, Prajapati, Beniwal, Singh. Supports language identification, POS tagging, NER, matrix-language identification, translation, and text normalisation.
PagePaperGitHub