MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202541052436 A) filed by Manipal Academy Of Higher Education, Manipal, Karnataka, on May 30, 2025, for 'a system and method for detecting polysemy in news articles using large language models.'
Inventor(s) include Anusha; and Manjula Shenoy K.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "The current invention offers a system and approach for identifying polysemous words in news articles through the utilization of large language models (LLMs). Conventional NLP systems face difficulties with polysemy, particularly in brief texts such as news headlines, where the context is restricted. This innovation employs sophisticated models like BERT and SenseBERT to identify polysemous words and clarify their meanings according to context. The system additionally produces synthetic training data to strengthen model resilience and increase the correctness of semantic comprehension. This invention provides a notable enhancement in NLP tasks such as news classification, search engine optimization and content filtering by assessing the model's performance through precision, recall and confusion matrix metrics. This technology is especially beneficial for news aggregation services, AI-enhanced journalism and tools that need precise semantic understanding in context-restricted texts."
Disclaimer: Curated by HT Syndication.
