MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122978 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 5, 2025, for 'machine learning system for adaptive procedural game content generation.'
Inventor(s) include Dr. D. Ajitha; Aryan Thapliyal; Mohamed Arshad Z H; and Boddoju Sai Amrutha.
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 present disclosure provides a machine learning system for adaptive procedural game content generation. The system includes a data collection module for collecting player behavior data, a machine learning analysis module for analyzing data to identify behavioral patterns, and a content generation module for generating game content based on identified patterns. The system operates in real-time for dynamic content adaptation during gameplay. The system includes a preprocessing module for normalizing structured telemetry data and tokenizing unstructured data using natural language processing, and a clustering module for segmenting players into behavioral archetypes. The machine learning analysis module includes reinforcement learning using proximal policy optimization, transfer learning between game genres, and explainable artificial intelligence using SHapley Additive exPlanations. The content generation module uses Binary Space Partitioning algorithms for structured layouts and Random Walk algorithms for organic paths."
Disclaimer: Curated by HT Syndication.
