MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202621092288 A) filed by Symbiosis International Deemed University on July 29, 2026, for Explainable Artificial Intelligence Framework For Transparent Real Estate Valuation.
Inventors include Raghav Metange; Vedant Shiraskar; Mahek Ghodmare; and Dr. Smita Singh Nirkhi.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: ABSTRACT EXPLAINABLE ARTIFICIAL INTELLIGENCE FRAMEWORK FOR TRANSPARENT REAL ESTATE VALUATION The present invention discloses an explainable artificial intelligence framework (100) for transparent real estate property valuation comprising a data preprocessing pipeline (200) that applies quantile-based outlier filtering, logarithmic target transformation, frequency-threshold locality grouping, one-hot encoding, and Z-score normalization to multi-city property listing data; a machine learning prediction engine (300) comprising a Random Forest Regressor selected based on achieving the lowest root mean squared error in real currency space; a SHAP explainability engine (400) that computes exact Shapley values using TreeExplainer and returns ranked feature attributions with monetary magnitudes and directional indicators; a backend API module (500) for serving predictions and explanations; a database module (600) for prediction logging and geocoded locality statistics; and a frontend user interface module (700) with an attribution panel (720) and a geospatial map component (730) for interactive locality-level price visualization. The framework enables transparent and interpretable property valuations in real-time. [
Disclaimer: Curated by HT Syndication.
