Biometric Keystroke Analysis for Exam Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for confirming the identity of examinees in computer-based essay assessments are inadequate in detecting fraudulent activities, such as identity impersonation, due to the large number of test takers and testing sites, making it difficult to accurately verify individual identities.
Innovation Solution
The use of biometric keystroke measure data analysis, involving feature extraction and similarity determination using direct distance or machine learning approaches, to identify potential identity mismatches by comparing keystroke patterns of examinees, with features derived from writing and digraph processes, and employing gradient boosted decision trees for enhanced accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional identity verification methods are used for examinees, then the verification process is simple to implement, but the detection accuracy of fraudulent activities is insufficient due to the large number of test takers and testing sites
Solution Approach 1:
The patent replaces conventional mechanical identity verification methods (such as ID checking and manual supervision) with biometric keystroke analysis. The system captures keystroke dynamics data during essay writing and uses machine learning models to automatically verify examinee identity, substituting physical verification processes with automated computational analysis that can accurately detect identity impersonation across large numbers of test takers and testing sites
Solution Approach 2:
The system enables self-service identity verification by having examinees naturally write essays during the verification process. The keystroke patterns generated during normal essay writing are automatically analyzed by the system to confirm identity, eliminating the need for separate verification procedures or additional user actions beyond the required essay writing task
2Reliability
If biometric keystroke measure data analysis is used to confirm examinee identity, then the detection accuracy of identity impersonation is improved, but the computational resource consumption increases
Solution Approach 1:
The system extracts only the essential biometric features from keystroke data that are most indicative of identity verification. Rather than analyzing all keystroke data in detail, the system identifies and extracts key temporal and dynamic features (such as keystroke timing, pressure patterns, and typing rhythm) that are sufficient for reliable identity confirmation, thereby reducing computational resource requirements while maintaining high detection accuracy
Solution Approach 2:
The system uses partial action by analyzing only the essay writing portion of the test rather than monitoring all test sections. The keystroke biometric analysis is applied selectively during the essay writing task where sufficient data can be captured, rather than continuously monitoring all test activities, thereby reducing overall computational resource consumption while still achieving reliable fraud detection
Data Source
AI summary
Biometric keystroke measure data derived from a computer-implemented long form examination taken by an examinee is received. Features are the extracted from the biometric keystroke measure data for the examinee. A similarity value is then determined, using one or more of a direct distance approach or a machine learning approach, for the extracted features relative to features extracted from biometric keystroke measure data derived from each of a plurality of other examinees while taking the long form examination. At least one of the determined similarity values is then identified having a value above a pre-defined threshold. The pre-defined threshold indicates a likelihood of the examinee being the same as one of the other examinees. Data can then be provided that characterizes the identification. Related apparatus, systems, techniques and articles are also described.