Bio-behavioral Authentication System for Digital Transaction Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital transactions are often compromised by unauthorized parties or malicious actors, leading to fraudulent activities due to stolen credentials or compromised accounts, especially in environments affected by Denial of Service (DoS) attacks, making it difficult to authenticate legitimate users and detect abnormalities in transaction risk.
Innovation Solution
A system utilizing AI/ML capabilities in a smart data hub and risk engine to continuously capture and analyze contextual and behavioral factors of user entities, developing bio-behavioral models to compare transaction requests against allocentric and egocentric factors, and determining risk scores to authenticate and authorize transactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional authentication methods are used, then ease of operation is maintained, but reliability of transaction security deteriorates due to stolen credentials and unauthorized access
Solution Approach 1:
The system performs preliminary actions by continuously capturing and analyzing contextual and behavioral factors before transactions occur, developing bio-behavioral models in advance to establish a baseline of normal user behavior patterns, which enables proactive risk assessment rather than reactive security measures
Solution Approach 2:
The system implements dynamic risk scoring by continuously updating bio-behavioral models with recent behavioral data and adjusting authentication requirements in real-time based on transaction risk levels, transitioning from static credential verification to dynamic behavioral analysis that adapts to changing threat landscapes
2Reliability
If bio-behavioral modeling with AI/ML is implemented, then reliability of fraud detection is improved, but device complexity increases due to multiple processors and continuous data capture
Solution Approach 1:
The system segments the security architecture into distinct functional components: a smart data hub for continuous behavioral data capture and processing, a risk engine for AI/ML-based analysis, and integration layers for transaction monitoring, allowing each component to be optimized independently and deployed in a modular fashion that manages complexity
Solution Approach 2:
The bio-behavioral modeling system serves multiple functions simultaneously: it authenticates user identity, detects fraudulent transactions, monitors network security, and provides risk scoring for authorization decisions, consolidating what could be separate security systems into a unified multi-functional platform
3Measurement precision
If continuous behavioral data capture is performed, then measurement precision of user anomalies is improved, but loss of time for data processing increases
Solution Approach 1:
The system applies partial action by focusing data capture and analysis efforts on the most discriminative behavioral factors and contextual elements that provide the highest signal-to-noise ratio for anomaly detection, rather than processing all possible data uniformly, enabling precise anomaly detection with reduced computational overhead
Solution Approach 2:
The system implements feedback loops where risk scores and anomaly detections from recent transactions are fed back into the bio-behavioral models to continuously refine behavioral baselines, allowing the system to adapt to legitimate behavior changes while maintaining sensitivity to fraudulent patterns, improving precision over time without increasing processing burden
Data Source
AI summary
Aspects of the disclosure provide techniques for using bio-behavior based information for providing and restricting access to a secure website or computer network and its assets to a user entity. The bio-behavior system and method 100 uses processes to learn models that relate heterogeneous data that connects the analog, physical space with the online/cyber world. The process is a nonparametric, probabilistic mixture model. The system 100 is capable of detecting behavioral patterns in mixed data composed of inputs of varying complexity. This includes the low-level, mainly unprocessed data generated by a user entity device's intrinsic sensors that monitor the internal state of the phone as well as extrinsic sensors that capture the state of the surrounding environment.