Encrypted Document Plagiarism Detection via Blockchain Tokens
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack an effective method for detecting plagiarism in encrypted documents without decrypting the content, which is essential for maintaining document security and privacy while ensuring accurate similarity analysis.
Innovation Solution
A blockchain-based system that utilizes encrypted tokens and frequency values, stored on a decentralized ledger, allows for the comparison of encrypted documents to determine similarity without decrypting the content, using techniques like message locked encryption and order preserving encryption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If document content is decrypted for plagiarism detection, then similarity analysis accuracy is improved, but document security and privacy are compromised
Solution Approach 1:
The patent introduces encrypted tokens as an intermediary representation of document content. These tokens capture semantic information while remaining encrypted, allowing similarity comparison without exposing the actual document content. The tokens act as a mediator between the need for plagiarism detection and the requirement for maintaining document confidentiality.
Solution Approach 2:
The patent transforms document content into a different parameter space - from raw text to encrypted tokens with associated frequency values. This parameter transformation enables mathematical operations for similarity detection while preserving the confidentiality of the original content through encryption.
2Object-affected harmful factors
If encrypted tokens are used for plagiarism detection, then document security is maintained, but the ability to detect plagiarism accurately deteriorates
Solution Approach 1:
The patent performs preliminary transformation of document content into encrypted tokens and frequency values before the plagiarism detection process. This preliminary action structures the encrypted data in a way that enables accurate similarity comparison through mathematical operations on the token frequencies, ensuring both security and detection accuracy from the outset.
Solution Approach 2:
The patent replaces the traditional mechanical approach of decrypting and comparing text with a mathematical approach operating on encrypted tokens. By substituting text-based comparison with mathematical operations on encrypted frequency vectors, the system maintains security while achieving accurate plagiarism detection through cosine similarity or other mathematical metrics.
3Measurement precision
If traditional plagiarism detection methods are used, then plagiarism can be detected accurately, but document privacy must be compromised through decryption
Solution Approach 1:
The patent extracts only the essential features needed for plagiarism detection - token identities and frequency values - while leaving the actual document content encrypted and confidential. This extraction approach isolates the minimal necessary information for comparison, maintaining document privacy while enabling accurate detection through frequency-based similarity analysis.
Data Source
AI summary
An example operation may include one or more of receiving a request to verify a first encrypted document from a computing device, retrieving a second set of encrypted tokens of a second encrypted document from a blockchain, determining a similarity value of the first encrypted document with respect to the second encrypted document based on a first set of encrypted tokens in the first encrypted document and the second set of encrypted tokens in the second encrypted document, and outputting the determined similarity value to the computing device in response to the request.