MUMBAI, India, Jan. 8 -- Intellectual Property India has published a patent application (202511104261 A) filed by Lovely Professional University, Phagwara, Punjab, on Oct. 29, 2025, for 'system for predicting fragrance odors.'
Inventor(s) include Dr. Arun Malik; Ginni Nijhawan; Dr. Alok Jain; and Monica Gulati.
The application for the patent was published on Dec. 12, under issue no. 50/2025.
According to the abstract released by the Intellectual Property India: "A system for predicting fragrance odors comprises a molecular dataset module 101 stores a curated dataset of 8,681 fragrance-related compounds with sensory annotations, a descriptor extraction module 102 derives functional group fingerprints, classical molecular descriptors, and Morgan structural fingerprints to represent molecular structures, the descriptors are processed in the machine learning model module 103, which employs frameworks such as Random Forest, XGBoost, or LightGBM, preferably utilizing an XGBoost model based on Morgan fingerprints for high accuracy, model performance is assessed by the evaluation and benchmarking module 104 using AUROC and AUPRC metrics, an interpretability and visualization module 105 generates continuous scent space representations aligning molecular and perceptual relationships, a predictive output module provides detailed fragrance odor predictions and chemical insights for fragrance design and research applications."
Disclaimer: Curated by HT Syndication.
