Histogram String Matching for Cyber Security Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Levenshtein distance calculations are computationally expensive and difficult to integrate with machine learning and data analytics due to their non-numeric nature, making string comparisons inefficient in computing environments, especially in authentication and security contexts.
Innovation Solution
Transforming strings into histogram features, which are numeric representations of character frequencies, allowing for fast distance computations and easier application of machine learning algorithms, and determining an acceptable histogram threshold based on true and false positive rates for approximate string matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Levenshtein distance calculations are used for string comparison, then measurement precision is improved, but productivity deteriorates due to computational expense
Solution Approach 1:
The patent transforms string data from categorical form to numerical histogram form, changing the parameter representation. This allows the use of efficient numerical distance calculations (Euclidean distance on histograms) instead of computationally expensive Levenshtein distance calculations on strings, while maintaining acceptable comparison accuracy through the histogram abstraction that captures character frequency patterns.
2Measurement precision
If Levenshtein distance is used for string matching, then measurement precision is improved, but loss of time increases due to computational complexity
Solution Approach 1:
The patent creates a histogram copy or representation of the original string data. Instead of directly comparing strings using Levenshtein distance, the system computes histograms that capture the essential character frequency information and compares these histogram representations using faster numerical methods, reducing calculation time while preserving measurement precision.
3Productivity
If histogram thresholding is applied for approximate matching, then productivity is improved through faster comparisons, but measurement precision deteriorates
Solution Approach 1:
The patent applies histogram thresholding as a partial action that provides approximate matching rather than exact matching. By using distance thresholds on histogram comparisons, the system achieves fast productivity-improving comparisons that are sufficient for many security applications where approximate matching (allowing for minor variations) is acceptable, balancing speed and precision appropriately for the use case.
Data Source
AI summary
A method for performing approximate string matching comprises: obtaining a predefined distance error threshold and user information, the user information including a list of strings; calculating distance metrics for pairs of strings in the list of strings; determining a positive list and a negative list based on the calculated distance metrics for the pairs of strings; determining histogram representations of the list of strings; determining true positive rates and false positive rates for one or more histogram thresholds based on the positive list, the negative list, and calculated histogram distances between pairs within the histogram representations; and obtaining an acceptable histogram threshold from the one or more histogram thresholds based on an acceptable true positive rate and an acceptable false positive rate.