Confidence Broker Normalizes Identity Authentication Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing identity authentication systems are vulnerable to misuse and theft due to their binary nature, which does not account for varying levels of confidence in identity credentials, leading to potential security compromises when insiders or stolen credentials are used outside their intended geographic boundaries.
Innovation Solution
A Confidence Broker system that normalizes confidence indications from different analytics systems and maps them to policy enforcement systems, using scaling factors and offsets to create a common range for comparison, and mediates conflicting confidence values to provide consistent consumer-specific confidence values, preventing oscillations and enhancing security interoperability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If binary identity authentication is used, then the authentication process is simple and fast, but the system is vulnerable to misuse and theft of credentials
Solution Approach 1:
The patent transforms the binary authentication result (valid/invalid) into a continuous confidence score that reflects the likelihood of legitimate use. This parameter change allows the system to distinguish between stolen credentials used outside their intended context and legitimate credentials, thereby improving security robustness while maintaining authentication speed.
Solution Approach 2:
The patent introduces an intermediary confidence scoring mechanism between the binary authentication result and the final access decision. This intermediary layer analyzes additional factors (geographic location, device characteristics, behavior patterns) to generate a confidence score, which then guides the access decision. This resolves the contradiction by adding security analysis without significantly impacting authentication speed.
2Adaptability or versatility
If multiple analytics systems with different confidence scales are integrated, then comprehensive confidence evaluation is achieved, but normalization and comparison become complex
Solution Approach 1:
The patent creates a universal confidence score framework that can accommodate multiple analytics systems with different confidence scales. By defining a standardized confidence score range (0-100) and implementing normalization transformations for each system, the framework enables comprehensive multi-system integration without requiring complex custom handling for each system's unique scale.
Solution Approach 2:
The patent applies parameter transformation techniques to convert confidence scores from different analytics systems into a common standardized scale. Each analytics system's confidence output is transformed using system-specific normalization parameters (linear transformation with slope and intercept) to produce comparable standardized confidence scores, thereby simplifying integration while maintaining comprehensive evaluation capabilities.
3Adaptability or versatility
If confidence values are normalized to a common range, then comparison across systems is enabled, but information loss may occur during transformation
Solution Approach 1:
The patent employs reversible linear transformation (normalization) to convert confidence scores to a common range while preserving the relative relationships between different confidence levels. The transformation uses system-specific slope and intercept parameters that maintain the original confidence distribution's shape and relative spacing, minimizing information loss while enabling cross-system comparison.
Solution Approach 2:
The patent incorporates feedback mechanisms that track the original confidence values alongside the normalized scores. This allows the system to reference the original nuanced confidence measurements when making decisions, preventing information loss from affecting the final outcome while still benefiting from the comparability enabled by normalization.
4Reliability
If mediation is implemented to resolve conflicting confidence values, then consistent policy enforcement is achieved, but additional processing time is required
Solution Approach 1:
The patent implements preliminary confidence aggregation and mediation before the final policy enforcement decision. By pre-processing conflicting confidence values from multiple analytics systems and establishing a mediated confidence score in advance, the system avoids time-consuming conflict resolution during critical decision moments, thereby maintaining policy enforcement consistency while minimizing additional processing time.
Solution Approach 2:
The patent transforms multiple conflicting confidence parameters into a single mediated confidence score through weighted aggregation. This parameter reduction consolidates multiple confidence measurements into one comprehensive value that reflects the overall legitimacy assessment, enabling consistent policy enforcement without requiring complex real-time conflict resolution and reducing processing time.
Data Source
AI summary
A Confidence Broker System is disclosed. One embodiment of the present invention includes a confidence broker (10) which communicates with a plurality of confidence producers (12A, 12B, 12C) and a plurality of confidence consumers (14A, 14B, 14C). Communications between these elements is conducted via a communications infrastructure (16). The confidence broker (10) also includes a communications interface (42) which is connected to a protocol converter (44). The protocol converter (44) is connected to a confidence normalizer (46). The confidence normalizer (46) is connected to a confidence mediator (48). The confidence mediator (48) is connected to a confidence mapper (50). The confidence mapper (50) is connected to the protocol converter (44). Each of the protocol converter (44), the confidence normalizer (46), the confidence mediator (48) and the confidence mapper (50) is connected to a storage device (52).