MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202621057725 A) filed by Sage University on May 06, 2026, for System And Method For Detecting Fabricated Digital Content.
Inventors include Dr. Prashant Jain; Dr. Ati Jain; Ajeet Pal; Abhay Gaur; and Krishna Goyal.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: ABSTRACT The present disclosure relates to a system (100) and a method for detecting fabricated digital content using textual feature analysis. The system (100) includes a text segmentation circuit (102) that decomposes textual content into a headline portion, a body portion, and a plurality of sentence windows. A feature extraction circuit (104) extracts sentiment features, certainty language measures, and factual density measures. A consistency analysis circuit (106) performs cross-layer analysis to determine an intra- document divergence metric set by determining a sentiment divergence value, generating an emotional intensity variation signal, and detecting a certainty-to-factual-density mismatch. A scoring circuit (110) generates a probabilistic confidence score (202) exclusively from the intra-document divergence metric set and feature influence weights from an adaptive weight register (112). The probabilistic confidence score (202) is delivered as an authenticity classification signal to a content moderation pipeline (204). FIG. 1
Disclaimer: Curated by HT Syndication.
