Machine Learning for Thesis Image and Associated-Text Plagiarism Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing plagiarism detection techniques fail to accurately detect plagiarism related to images and text associated with images in a thesis, such as sentences referring to the images.
Innovation Solution
A method and system using machine learning models to extract and analyze images and text from a thesis, determining similarities and types of plagiarism by applying feature data to multiple models, including a first model for detecting plagiarism and a third model for identifying plagiarism types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional plagiarism detection programs are used, then text plagiarism can be detected, but image plagiarism and associated text plagiarism cannot be detected
Solution Approach 1:
The system employs multiple machine learning models (first ML model for image feature extraction, second ML model for text feature extraction, third ML model for plagiarism type classification) that can handle different types of data (images and text) and perform multiple functions (feature extraction, similarity comparison, plagiarism detection, and classification), making the system versatile and adaptable to various plagiarism scenarios
Solution Approach 2:
The plagiarism detection process is divided into distinct segments: image processing segment (first ML model), text processing segment (second ML model), similarity comparison segment, and classification segment (third ML model). Each segment handles specific tasks independently, allowing the system to detect both image and text plagiarism with high precision
2Measurement precision
If multiple machine learning models are applied to detect both image and text plagiarism, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The complex detection task is segmented into specialized sub-tasks handled by different ML models: the first ML model processes images, the second ML model processes text, and the third ML model classifies plagiarism types. This segmentation allows each model to focus on specific functions, improving overall accuracy while managing complexity through modular design
Solution Approach 2:
Feature data serves as an intermediary between the input theses and the final plagiarism determination. The first and second ML models extract feature data from images and text respectively, which then serves as input for similarity comparison and subsequent classification by the third ML model, simplifying the overall processing pipeline
Data Source
AI summary
According to an aspect of the present disclosure, there is provided a thesis plagiarism detection method performed by a computing system. The thesis plagiarism detection method may comprise acquiring figure data for a target thesis, the figure data including images and text, acquiring first feature data for the figure data by applying the figure data to a first machine learning model, determining whether a thesis associated with second feature data having a similarity above a predetermined threshold with the acquired first feature data is found and determining the target thesis as a plagiarized thesis when it is determined that the thesis associated with the second feature data is found.