Typing Pattern User Identification via Touch Coordinates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying users in computerized systems, such as social media and payment systems, face challenges with high false positive rates and lack of real-time verification, especially in cross-border transactions, and require complex hardware and software, making them inconvenient and prone to misuse.
Innovation Solution
A method that uses touch coordinates from a touch-sensitive input interface to generate a unique typing pattern for each user, analyzed by algorithms like neural networks, to distinguish authorized from unauthorized users without requiring non-secret patterns, allowing for discreet and reliable real-time verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing risk-engine strategies are used for fraud detection, then fraud prevention is provided, but false positive rates are high causing inconveniences to cardholders and merchants
Solution Approach 1:
The patent changes the parameters used for fraud detection from traditional risk-engine strategies to behavioral biometric parameters including typing rhythm, key press duration, key travel time, and typing pressure. These new parameters enable more accurate user identification without increasing false positives, thereby maintaining fraud prevention reliability while improving user convenience.
Solution Approach 2:
The patent replaces the mechanical risk-engine decision-making system with a behavioral biometric analysis system that continuously monitors and analyzes typing behavior patterns. This substitution enables more precise user identification based on unique behavioral characteristics rather than rule-based risk scoring, reducing false positives while maintaining security.
2Measurement precision
If complicated software and specific hardware components are used for behavioral analysis, then user identification accuracy is improved, but device complexity increases
Solution Approach 1:
The patent implements self-service by utilizing the existing touchscreen interface and built-in sensors of standard devices to collect behavioral data. The system automatically monitors typing behavior without requiring external hardware or complex software installations, thereby achieving high identification accuracy while maintaining device simplicity.
Solution Approach 2:
The patent makes the behavioral biometric system universal by designing it to work on standard touchscreen devices with basic sensors already present in modern smartphones and tablets. The solution leverages existing device capabilities for multiple purposes including typing analysis, pressure sensing, and temporal measurement, eliminating the need for specialized hardware while maintaining high identification accuracy.
3Ease of operation
If non-secret patterns are presented to users for authentication, then identification can be performed, but users may deliberately change their behavior making the method prone to misuse
Solution Approach 1:
The patent applies preliminary action by continuously collecting and analyzing behavioral data during normal typing activities before authentication is required. The system establishes baseline behavioral patterns during regular device usage, enabling seamless authentication without interrupting the user or requiring deliberate pattern reproduction, thereby maintaining both convenience and security.
Solution Approach 2:
The patent ensures continuity of useful action by continuously monitoring behavioral parameters throughout device usage rather than requiring intermittent authentication actions. The typing behavior analysis occurs naturally during normal interaction with the device, providing continuous authentication verification without disrupting the user experience or creating opportunities for deliberate behavior manipulation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
One or more users (42) are identified based on typing pattern and/or behavior tn a computerized system. A typing input is generated by one or more authorized users (42) by touching a touch sensitive input interface (12) at a touch coordinate that is linked to a given interface element displayed by the input interface (12). Collected touch coordinates are processed to generate a typing pattern specific of the authorized users (42), and the user-specific typing pattern is stored in a database (20). Touch coordinates generated by an un-identified user are processed to generate a typing pattern specific of the un-identifled user (42), and the typing pattern of the un-identified user (42) is compared with the user-specific typing pattern stored in the database (20). The un-identified user is identified as (42) as one of the authorized users or as a non-authorized user.