Intellectual Property India Publishes Patent Application for “an Artificial Intelligence-Based System For Content-Aware Classification And Selective Deletion Of Digital Files For Storage Optimization”

MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641074967 A) filed by Vellore Institute Of Technology on June 17, 2026, for “an Artificial Intelligence-Based System For Content-Aware Classification And Selective Deletion Of Digital Files For Storage Optimization”.

Inventors include Dr. Kumaresan P; Chandra Pravesh; and Dr. Yokesh Babu Sundaresan.

The application for the patent was published on June 26, 2026, under issue no. 26/2026.

Abstract: The present invention relates to an artificial intelligence-based system and method for content-aware classification and selective deletion of digital files for storage optimization. The system comprises a storage scanning module configured to acquire digital files from one or more storage repositories, a user preference management module configured to receive user-defined retention categories, a multi-modal analysis engine comprising media, audio, document, and mail analysis modules, a semantic classification engine, a staging repository, a review portal, and a deletion management module. The multi-modal analysis engine utilizes computer vision, action-recognition, acoustic analysis, optical character recognition, semantic embedding, and mail-content analysis techniques to evaluate the substantive content of digital files including images, videos, audio files, documents, and electronic mail messages. The semantic classification engine determines content relevance with respect to user-defined categories and classifies files as required or not-required. Files classified as not-required are transferred to a staging repository and presented through a review portal that enables user validation and selective restoration prior to deletion. Following completion of the review process, remaining not-required files are permanently removed to reclaim storage space. The system further supports custom classification categories, attachment-aware email analysis, storage-recovery estimation, duplicate-content identification, and storage-tier recommendations. By performing semantic content evaluation instead of metadata-based analysis, the invention enables intelligent identification of valuable digital assets, reduction of digital clutter, optimization of storage utilization, and preservation of user control during automated file lifecycle management. Fig 1 to 11.

Disclaimer: Curated by HT Syndication.