MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202621044150 A) filed by Prof. Supriya S. Gorde; Ayush Dinesh Deshmukh; Karan Ram Bainade; Bhagyashri Jibhau Ahire; and Tanaya Dhondiraj Arvikar on April 07, 2026, for Explainable Artificial Intelligence Based Fairness-Aware Scholarship Recommendation System Using Hybrid Machine Learning And Calibrated Match Probability Scoring.
Inventors include Prof. Supriya S. Gorde; Ayush Dinesh Deshmukh; Karan Ram Bainade; Bhagyashri Jibhau Ahire; and Tanaya Dhondiraj Arvikar.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: The current invention describes a computer-implemented scholarship recommendation system that uses hybrid similarity computation between scholarship eligibility vectors and student feature vectors to produce calibrated eligibility probability outputs. Natural language processing techniques are used to extract structured eligibility criteria from scholarship descriptions. These elements are then merged with encoded student profile information in a hybrid recommendation engine that uses collaborative filtering prediction and content-based similarity. Logistic score calibration is used to convert similarity scores into comprehensible eligibility probability outputs. To offer organized eligibility reasoning, explainability approaches are used to provide feature-level contribution values that correlate to individual student traits. To lessen prejudice among applicant groups, a fairness adjustment module applies demographic parity limits to recommendation results. The approach offers a comprehensive framework for producing explainable, fairness-conscious, and probability-based scholarship eligibility recommendations within a computer- implemented decision-support environment.
Disclaimer: Curated by HT Syndication.
