Intellectual Property India Publishes Patent Application for Explainable Artificial Intelligence-Based Framework For Real-Time Decision Support And Adaptive Decision Generation

MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641098997 A) filed by Dr. R. Catherin Ida Shylu; Dr. R. Rajpriya; Dr. V. Veeramanikandan; Dr. K. Kavitha; Dr. M. Jeyakarthic; and Mrs. R. Suganya on August 16, 2026, for Explainable Artificial Intelligence-Based Framework For Real-Time Decision Support And Adaptive Decision Generation.

Inventors include Dr. R. Catherin Ida Shylu; Dr. R. Rajpriya; Dr. V. Veeramanikandan; Dr. K. Kavitha; Dr. M. Jeyakarthic; and Mrs. R. Suganya.

The application for the patent was published on August 28, 2026, under issue no. 35/2026.

Abstract: [088] The present invention relates to an explainable artificial intelligence-based framework for real-time decision support and adaptive decision generation. The framework comprises a real-time data acquisition module, data normalization and quality assessment module, feature engineering module, artificial intelligence decision engine, model monitoring module, explainability orchestration engine, explanation validation module, confidence assessment engine, decision policy engine, decision presentation interface and provenance repository. The framework receives heterogeneous real-time data, evaluates data quality, generates contextual features and produces a decision using one or more artificial intelligence models. An explainability orchestration engine dynamically selects an explanation technique according to model architecture, decision type, operating condition, computational latency, decision criticality and user requirements. The generated explanation is validated for fidelity, stability, consistency and sensitivity, while decision reliability is determined using model confidence, input-data quality and explanation reliability. The framework selectively permits automated action, operator-assisted action, human verification or additional data acquisition based on the determined reliability. A provenance repository maintains an association between the input state, model version, decision, explanation and reliability information, thereby providing a technically integrated mechanism for transparent, traceable and adaptive real-time artificial intelligence-based decision support. Accompanied Drawing [FIGS. 1-2]

Disclaimer: Curated by HT Syndication.