Behavioral Biometrics for Human-Machine Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing authentication mechanisms on computing devices are vulnerable to security threats due to malicious programs and bots, with conventional anti-virus/malware detection software having limitations and users often not employing the best protection, leading to unauthorized access and attacks like brute force, bot/botnet, man-in-the-middle, and man-in-the-browser attacks.
Innovation Solution
A system and method that differentiate human users from malicious non-human users by analyzing behavioral characteristics and leveraging cognitive differences, using processing logic to collect and analyze keystroke, touch, mouse, and sensor data, and employing UI transformations and seeded captchas to prevent malicious code from interpreting user interfaces, thereby distinguishing human from non-human users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional anti-virus/malware detection software is used, then malware detection capability is provided, but detection accuracy and reliability are insufficient due to security weaknesses and human errors
Solution Approach 1:
The patent replaces conventional anti-virus/malware detection software with a behavioral analysis system that monitors user interactions, keystroke dynamics, mouse movements, and sensor data patterns. This substitution moves from traditional malware signature matching to real-time behavioral biometrics, enabling more accurate differentiation between human and non-human users.
Solution Approach 2:
The patent introduces an intermediary behavioral analysis layer between the user interface and the authentication system. This intermediary collects and analyzes behavioral characteristics (keystroke patterns, touch pressure, mouse trajectories, sensor data) to create a behavioral profile, which then informs authentication decisions. This intermediary layer adds detection capability without requiring users to install or configure anti-virus software.
2Reliability
If authentication mechanisms are implemented, then unauthorized access is prevented, but security threats persist due to malicious programs and bots
Solution Approach 1:
The patent implements continuous feedback loops where the system monitors behavioral characteristics in real-time and adjusts authentication decisions accordingly. The system compares current behavioral patterns against established human user profiles and updates its assessment dynamically, allowing it to detect deviations that indicate bot or malware activity during ongoing interactions.
Solution Approach 2:
The patent creates a universal behavioral analysis framework that works across multiple authentication scenarios and device types (mobile and non-mobile). The same core principles of behavioral monitoring apply to various attack vectors including brute force, bot/botnet, man-in-the-middle, and man-in-the-browser attacks, making the security solution broadly applicable despite the diversity of threat vectors.
3Device complexity
If users do not employ the best anti-virus/malware detection software, then device simplicity is maintained, but security vulnerabilities increase
Solution Approach 1:
The patent enables the system to perform self-service security monitoring by automatically collecting, analyzing, and acting on behavioral data without requiring user intervention or configuration. The system autonomously establishes baseline behavioral profiles, detects anomalies, and triggers appropriate security responses, freeing users from the need to manage complex anti-virus software while maintaining strong security protection.
Data Source
AI summary
A method including collecting, by a processing device, raw data regarding an input to fill a form field. The method further includes converting, by the processing device, the raw data to test data, wherein the test represents behavioral characteristics of the entry of the input. The method further includes identifying a human characteristic model corresponding to the behavior characteristics of the entry of the input. The method further includes generating a predictor from a comparison of the test data against the corresponding human characteristic model. The predictor includes a score indicating a probability that the input originated from a human user or from a malicious code imitating the human user.