MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641049682 A) filed by Anurag University, Hyderabad, Telangana, on April 18, for 'ai-based multilingual context-aware hate speech detection system for social media content analysis.'
Inventor(s) include Dr. Purnachary Munigadapa; Dr. Ravinder Reddy Bukkasamudram; Dr. Pocha Rajasekhar Reddy; Dr. Chokkam Dastagiraiah; and Dr. Gunda Vishnu Murthy.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses an Artificial Intelligence-based multilingual context-aware hate speech detection system. The system (100) comprises a data acquisition module (104) configured to collect social media content (102), followed by a multilingual preprocessing module (106) for language identification and text normalization. A contextual feature extraction engine (108) generates semantic embeddings using transformer-based models, which are processed by a classification engine to categorize content into hate speech, offensive, or neutral classes (110). A context-aware conversation analysis module (112) refines the classification by analyzing surrounding textual interactions to detect implicit (114) and context-dependent hate speech. The system further includes a moderation interface for presenting results and a data storage layer for maintaining processed information. An adaptive learning module (116) continuously updates the model using feedback, thereby improving detection accuracy. The invention provides a scalable and efficient framework for real-time multilingual hate speech detection with enhanced contextual understanding."
Disclaimer: Curated by HT Syndication.
