MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641087023 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on July 16, 2026, for A Hybrid Neural Machine Translation And Topic Modeling Framework For Multilingual Social Media Trend Discovery And Content Analysis.
Inventor includes Dr. Kriti Ohri.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: Social media sites have transformed communication and content sharing; however, India's linguistic diversity is a huge challenge to automatic content identification and classification. Conventional trend detection techniques, including hashtag-based analysis, tend to provide fragmented information and limited visibility. Language barriers also hamper user interaction and smooth interaction across geographies. To solve these challenges, this research suggests a hybrid system that combines rule- based methods with neural machine translation (NMT) for efficient multilingual language detection and normalization. For content comprehension and trend discovery, we use and compare two sophisticated topic modeling methods: Latent Dirichlet Allocation (LDA) and BERTopic, which is founded on BERT embeddings. Whereas LDA provides interpretable and probabilistic topic models, BERTopic uses contextual embeddings for better coherence and semantic depth. A comparative study of the two models is done based on primary evaluation criteria like topic coherence, topic diversity, and topic coverage, emphasizing the strengths and trade-offs of both methods. The system also includes interactive visualizations and a user feedback loop to enhance topic relevance and precision. Our methodology closes linguistic differences, increases content prominence, and promotes social inclusion on social media sites, eventually making cross-language communication more effective and trends easier to find.
Disclaimer: Curated by HT Syndication.
