MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122935 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 5, 2025, for 'ai-powered cyberbullying detection system with multimodal content analysis.'
Inventor(s) include Dr. S. Bhavani; and Mr. Smaran Lal Shrestha.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a computer-implemented method for detecting cyberbullying content in digital communications. The method receives user-submitted content including textual data and image data, processes the textual data through a multi-stage analysis pipeline including pattern matching using regular expressions to identify harmful language patterns, keyword detection using a curated list of harmful terms, and sentiment analysis to calculate sentiment scores. The method processes the image data through optical character recognition to extract embedded text from the image data and object detection using a pre-trained neural network to identify harmful visual content. The method classifies the user-submitted content as cyberbullying based on detection results from the multi-stage analysis pipeline and the processing of the image data, and generates detection results with explanatory information indicating reasons for the classification."
Disclaimer: Curated by HT Syndication.
