Online Exam Cheating Detection via Screen Capture Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for detecting cheating in online tests lack effectiveness in identifying suspicious activities, such as copying text from examination windows to non-examination windows, due to limitations in screen capture and image analysis techniques.
Innovation Solution
A computer system and method that performs screen captures of user computers, extracts text and image data from non-examination application windows using image analysis, and assesses the relevance or matching with examination content to determine the likelihood of cheating by comparing substrings or using machine learning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional screen capture techniques are used, then the system can capture user screens, but it fails to accurately identify suspicious activities such as copying text between windows
Solution Approach 1:
The system segments the screen capture analysis by identifying and separating different application windows, then individually analyzing each window's content. This allows precise detection of text copying between examination and non-examination windows while maintaining manageable system complexity through modular processing of discrete window elements.
Solution Approach 2:
The system introduces an intermediary analysis layer that compares text content across different application windows. This intermediary text comparison mechanism detects suspicious copying activities without requiring complex direct monitoring of user actions, thereby improving detection accuracy while keeping the system architecture relatively simple.
2Reliability
If the system monitors all application windows, then it can detect cheating activities, but it increases processing time and computational resources
Solution Approach 1:
The system applies local quality analysis by focusing intensive text comparison and image analysis only on non-examination windows that are active or suspicious, rather than uniformly processing all windows. This selective approach maintains high cheating detection reliability while significantly reducing overall processing time and computational resource consumption.
Solution Approach 2:
The system performs partial analysis by monitoring key indicators such as window focus changes and active application states rather than continuously analyzing all screen content. This partial monitoring approach maintains sufficient detection reliability while minimizing time loss and computational overhead.
3Measurement precision
If the system performs detailed image analysis on all screens, then it can identify copied text and images, but it increases computational complexity and processing overhead
Solution Approach 1:
The system extracts and isolates specific elements of interest (text strings and images) from the full screen capture, then applies detailed image analysis only to these extracted elements in non-examination windows. This extraction approach enables precise identification of copied content while reducing image analysis complexity by avoiding processing of the entire screen and examination window content.
Data Source
AI summary
There is provided a computer system of detecting user cheating in an online test, the system comprising a processing circuitry configured to: obtain capability of performing screen capture of a user computer; perform a screen capture of the user computer; perform at least one of: i) extracting one or more text strings from one or more non-examination application windows of the userscreen image, obtaining at least part of an examination question text that has been presented in an application window on the user computer, and for at least one of the one or more extracted text strings, determining a degree of relevance to the examination question text, and ii) determining, using image analysis techniques, a degree of matching between at least part of the at least one userscreen image, and data associated with a suspicious application.