MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123341 A) filed by Vardhaman College Of Engineering, Hyderabad, Telangana, on Dec. 7, 2025, for 'neurosymbolic reasoning system for explainable ai in complex decision-making tasks.'
Inventor(s) include Mr. K. Anvesh; Dr. Sameera Khan; Mr. S Ranjith Reddy; Ms. T Prashanthi; Mr. S Satheesh Kumar; and Ms. P. Swetha.
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 invention provides a neurosymbolic reasoning system that integrates neural perception with symbolic logic to deliver explainable and context-aware decision-making in complex tasks. The system combines deep learning models for processing unstructured data with a symbolic knowledge graph and rule-based inference engine to produce decisions that are both accurate and interpretable. A neurosymbolic mapping interface ensures seamless conversion between neural embeddings and symbolic predicates, while a conflict-resolution unit manages inconsistencies between learned patterns and formal rules. A traceability engine generates human-understandable explanations by linking neural confidence scores, activated rules, and predicate evaluations. The invention enables transparent, reliable, and high-performance AI suitable for domains requiring accountability and logic-grounded reasoning."
Disclaimer: Curated by HT Syndication.
