Online Test Cheating Detection Using DNN Habit Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Proctoring online tests is challenging due to limited visual perception, making it difficult to detect cheating actions like looking at external materials, and existing automated systems based on eye gaze analysis are unreliable due to false alarms from normal behaviors.
Innovation Solution
A deep neural network (DNN) is used to monitor test taker behavior by combining pre-acquired habit data with real-time eye gaze and body movement data from camera inputs, distinguishing abnormal from normal behavior to detect cheating.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated systems monitor test takers' eye gaze to detect cheating, then cheating detection capability is improved, but false alarm rate increases due to normal thinking behaviors
Solution Approach 1:
The patent combines multiple monitoring dimensions (eye gaze direction, head position, shoulder movement, upper body posture) into a unified behavior analysis system. Instead of relying solely on eye gaze data, the system integrates data from multiple body parts to comprehensively determine cheating behavior, thereby reducing false alarms caused by normal thinking behaviors that may involve eye movement.
Solution Approach 2:
The patent introduces a pre-test phase as an intermediary step to establish each test taker's normal behavior patterns before the actual test. This baseline behavior data serves as a reference for comparing during the test, allowing the system to distinguish between normal thinking behaviors and actual cheating actions, thus reducing false alarms.
2Area of stationary object
If proctors monitor all learners with cameras, then visual coverage is improved, but detection accuracy decreases due to limited visual perception and inability to continuously attend to each learner
Solution Approach 1:
The patent replaces the mechanical human proctoring system with an automated computer-based monitoring system. The system uses cameras to capture visual data and employs algorithms to automatically analyze behavior patterns, eliminating the limitations of human visual perception and attention capacity while maintaining comprehensive monitoring coverage.
Solution Approach 2:
The patent transitions from two-dimensional visual observation to multi-dimensional behavior analysis by incorporating eye gaze direction, head position, shoulder movement, and upper body posture. This multi-dimensional approach enables more accurate cheating detection compared to traditional visual monitoring alone.
Data Source
AI summary
Embodiments of the invention are directed to a method for providing online test cheating detection. A non-limiting example of the method includes acquiring a test taker's normal test behavior, also referred to as habit data, through a pre-test. During the main test, the habit data is fed as one of inputs to the deep neural network (DNN) along with other real time inputs that represents eye gaze direction and movements of other body parts such as the head, shoulder, and upper body. These real-time input data are extracted from the visual data captured by the test taker's camera. The deep neural network (DNN) is pre-trained using a pre-existing database to distinguish abnormal or suspicious behavior from normal test behavior.