MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202611115954 A) filed by Panipat Institute Of Engineering And Technology Piet; Er. Rajeev Kumar; Dr. Sudhir Mahajan; Dr. Nisha; Kiran Gupta; and Rakesh Verma on September 28, 2026, for System And Method For Context-Aware Detection Of Anomalous Financial Transactions Using Adaptive Transaction Pattern Analysis.
Inventors include Er. Rajeev Kumar; Dr. Sudhir Mahajan; Dr. Nisha; Kiran Gupta; and Rakesh Verma.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: [044] A computer-implemented system and method are disclosed for context-aware detection of anomalous financial transactions using adaptive transaction pattern analysis. The system receives a transaction record and generates transaction-level features, retrieves an entity-specific behavioural profile, derives multi-scale temporal descriptors from preceding transaction events, and derives relationship descriptors from an interaction representation linking transaction entities. The descriptors are fused into a contextual transaction representation from which a plurality of deviation components is determined. A calibration engine applies context-dependent weighting and a dynamically maintained baseline to generate a calibrated anomaly indicator. A drift and adaptation controller selectively updates short-term and long- term behavioural states so that persistent legitimate behavioural changes can be incorporated while high-deviation observations are prevented from immediately contaminating the behavioural baseline. An evidence vector identifies behavioural, temporal or relational dimensions contributing to the anomaly indicator, and verified outcomes may be used to update calibration or adaptation parameters. Accompanied Drawing [FIGS. 1-2].
Disclaimer: Curated by HT Syndication.
