MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202611050845 A) filed by Kulvinder Singh; and Sanjeev Dhawan on April 21, 2026, for A Hybrid Model For Recommendation System Architecture Using Genetic & Boosting Algorithm.
Inventors include Kulvinder Singh; and Sanjeev Dhawan.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: The invention is a hybrid (the combination of two or more different architectures-data models) recommendation system architecture that incorporates a balance of data, feature optimization and ensemble learning strategies, to hit a significant enhancement of the precision of the recommendations, in particular with unbalanced data. It takes user-interaction data and item metadata to input information and pre-processes and extracts features before applying a Synthetic Minority Over-Sampling Technique (SMOTE) that generates artificial samples representing minority classes. It is based on a genetic algorithm to evolve and refine subsets of features over generations through a background of crossover, mutation and selection according to a fitness function. Increased features are combined to submit inputs into a boosting- based ensemble model to educate an ensemble of various feeble learners with variable weights to decrease inaccuracy in the classification. Hybrid structure delivers ranked or customized recommendations to the end users and facilitates real time refreshment upon receiving new data of interaction. The invention is more precise, recalls more and more credible than more conventional recommendation schemes and the invention can be used in large and distributed systems.
Disclaimer: Curated by HT Syndication.
