PDF Content Masking Attack Detection via Glyph-Text Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PDF document security is compromised by content masking attacks, where custom fonts are used to alter text without changing the visual appearance, allowing attackers to manipulate reviewer assignments, evade plagiarism detection, and subvert search engine indexing, as current systems rely on text extraction tools that cannot verify font authenticity.
Innovation Solution
A networked computing environment with a masking engine that includes a document handler, mask identifier, font mapper, and OCR engine to detect and thwart content masking attacks by constructing masking fonts and comparing character glyphs to text strings, and a font verification method to distinguish between legitimate and masked characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If text extraction tools are used to verify PDF content, then text extraction efficiency is improved, but detection precision of content masking attacks deteriorates because tools cannot verify font authenticity
Solution Approach 1:
The patent introduces an intermediary verification process that compares the visual glyph rendering against the extracted text string. This intermediary step acts as a mediator between the text extraction tool and the content verification system, enabling detection of font-based masking attacks without sacrificing extraction efficiency. The comparison mechanism serves as a bridge that allows standard text extraction tools to be enhanced with attack detection capability.
Solution Approach 2:
The system performs preliminary font verification by comparing glyph rendering with extracted text before final content acceptance. This preliminary action checks whether the visual representation matches the extracted string, preventing masked content from being accepted as legitimate. The verification happens in advance of final content processing, maintaining efficiency while ensuring accuracy.
2Ease of operation
If custom fonts are used to mask content, then ease of manipulation is improved, but document integrity deteriorates
Solution Approach 1:
The patent implements preliminary anti-action by pre-establishing verification rules that compare glyph rendering with extracted text strings. Before content manipulation can succeed, the system has already prepared detection mechanisms that will identify any discrepancies caused by font-based masking. This preliminary defensive action counteracts the ease of manipulation provided by custom fonts.
Solution Approach 2:
The system provides feedback by comparing the visual glyph output with the extracted text input and alerting when they do not match. This feedback loop allows the system to detect when custom fonts are used to mask content, thereby maintaining document integrity despite the ease of manipulation that custom fonts provide.
3Productivity
If automated reviewer assignment systems are used, then assignment efficiency is improved, but system reliability deteriorates due to vulnerability to content masking attacks
Solution Approach 1:
The patent applies preliminary action by implementing font verification checks before the automated reviewer assignment system processes documents. The system pre-verifies that the extracted text matches the visual rendering, preventing masked content from entering the assignment pipeline. This preliminary verification maintains system reliability while preserving assignment efficiency.
Solution Approach 2:
The verification mechanism serves as an intermediary between the PDF document and the automated reviewer assignment system. It filters out maliciously masked content before it can affect assignment decisions, thereby maintaining both the efficiency of automated processing and the reliability of assignment outcomes.
4Measurement precision
If plagiarism detection systems are used, then detection capability is improved, but vulnerability to masking attacks worsens due to inability to verify font authenticity
Solution Approach 1:
The patent implements feedback by comparing the extracted text with the visual glyph rendering and using this comparison to detect masking attacks. The feedback mechanism allows the plagiarism detection system to identify when fonts have been manipulated to hide plagiarized content, thereby maintaining detection capability while reducing vulnerability to masking attacks.
Solution Approach 2:
The glyph-text comparison mechanism acts as an intermediary verification layer between the plagiarism detection system and the document content. It detects font-based masking before the plagiarism detection algorithm processes the text, preventing masked plagiarized content from evading detection.
Data Source
AI summary
The embodiments present a new class of content masking defenses against the Portable Document Format (PDF) standard. The defenses can identify attacks that cause documents to appear different than the underlying content extracted from the documents. A content masking defense method can include identifying a content masking attack by scanning a document file to extract a character code of a character appearing in the file. Next, the character is rendered based on a font that is embedded in the document file. Optical character recognition can be performed on the rendering, and a content masking attack can be identified based on a comparison of a result of the optical character recognition against the character code of the character.