Automatic Hindi grammar tagging improves quantum language processing scalability
Scaling Hindi Quantum Natural Language Processing through Automatic Pregroup Supertagging
Computation and Language
Summary
Quantum Natural Language Processing (QNLP) tries to understand language by turning grammar rules into quantum circuits, but assigning these grammar labels by hand for Hindi is slow and limits progress. The authors studied how to automatically label Hindi words with their grammar roles using different computer models on a small set of sentences. They found that simpler models using word context worked best, reaching about 65% accuracy. This shows that automatic grammar tagging for Hindi is possible and can help expand QNLP tools to more languages without so much manual work.
What this means in practice
- •For natural language processing teams: Automatically assign detailed Hindi grammar types to speed up building quantum language models.
- •For multilingual ai developers: Integrate automatic Hindi grammar tagging to scale QNLP tools beyond English and reduce manual annotation needs.
Tested on one dataset.
Authors
Gautami Sanjay Naik, Krishna Bhatia, Mithun Paul Saint-Germain, H Aswath Babu
Abstract
Quantum Natural Language Processing (QNLP) uses pregroup grammars to translate grammatical structure into diagrammatic representations and quantum circuits. Recent Hindi QNLP work has shown that Hindi-specific pregroup grammars can support grammar-sensitive compositional models, but grammatical type assignment is still largely manual, limiting scalability. This paper formulates automatic Hindi pregroup supertagging as a token-level classification task. Using a manually annotated corpus of 380 Hindi sentences, we evaluate lexical, contextual, prompting-based, lexical-repair, and suffix/morphology-aware methods. Results show that simple lexical and contextual models are strong in this low-resource setting: contextual backoff achieves the best completed accuracy of 64.56\%, while raw Qwen2.5 prompting reaches only 11.65\%. Lexical repair raises LLM-assisted prediction to 64.08\%, demonstrating the value of constraining generative outputs with symbolic grammar knowledge. Diagnostic analysis further shows that seen and unambiguous tokens are much easier than unseen tokens, and suffix/morphology features improve karaka-token accuracy but not overall performance. These results show that automatic Hindi pregroup assignment is feasible and can reduce reliance on manual annotation in future multilingual QNLP pipelines.