Compression Signature Author Detection Reducing Feature Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for author detection across diverse Web content sources, such as blogs and user profiles, are costly and inefficient due to the complexity of feature extraction and scalability issues in uncontrolled environments, leading to poor detection rates.
Innovation Solution
A method using compression signatures to determine the similarity between documents from different sources, where documents with a compression distance below a threshold are considered authored by the same person, leveraging the Normalized Compressor Distance (NCD) and its variation (2NCD) to group documents by authorship.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If shallow parsing and machine learning methods are used for author detection, then detection capability is provided, but computational cost and complexity increase significantly
Solution Approach 1:
The patent extracts only the essential structural information from documents (such as HTML tag hierarchy, element nesting patterns, and document architecture) while discarding unnecessary content details. This selective extraction reduces computational complexity while maintaining sufficient features for accurate author detection.
Solution Approach 2:
The patent creates a simplified representation or copy of the document structure that captures authorial characteristics without requiring full document analysis. By working with condensed structural models rather than complete documents, the system reduces processing requirements while preserving detection accuracy.
2Reliability
If feature extraction is performed on all aspects of Web pages, then comprehensive author detection is achieved, but processing time and computational resources increase
Solution Approach 1:
The patent applies different levels of analysis to different parts of documents based on their importance for author detection. Structural elements that are more indicative of authorship (such as consistent heading patterns, navigation structures, and layout frameworks) receive detailed analysis, while less informative content is processed more lightly or skipped entirely.
Solution Approach 2:
The patent performs partial feature extraction, focusing only on the most discriminative structural characteristics needed for author detection. By using a subset of structural features rather than all possible features, the system achieves adequate detection accuracy with significantly reduced processing time.
3Reliability
If traditional author detection methods are applied to diverse Web content sources, then authorship can be identified, but scalability to large uncontrolled environments is poor
Solution Approach 1:
The patent develops a universal structural feature extraction framework that can handle multiple types of Web content sources (blogs, forums, social media, news sites) using the same core principles. The approach is domain-agnostic, allowing the system to adapt to new content types without requiring source-specific customization, thereby improving scalability.
Solution Approach 2:
The patent uses configurable parameters that can be adjusted for different content domains and detection requirements. By modifying extraction thresholds, feature weights, and similarity thresholds rather than changing the fundamental approach, the system can scale to different environments while maintaining detection effectiveness.
Data Source
AI summary
A method and system are provided for detection of authors across different types of information sources such as across documents on the Web. The method includes obtaining a compression signature for a document, and determining the similarity between compression signatures of two or more documents. If the similarity is greater than a threshold measure, the two or more documents are considered to be by the same author. Scored pairs of documents are clustered to provide a group of documents by the same author.The group of documents by the same author can be used for user profiling, noise reduction, contribution sizing, detecting fraudulent contributions, obtaining other search results by the same author, or mating a document with undisclosed authorship to a document of known author.