Authorship Analysis (In-Depth)
The In-Depth View of the Authorship analysis provides detailed insights into Metadata and AI Authorship.
1. Metadata Analysis
The Metadata Analysis offers detailed insights into the metadata of a document, helping to identify and flag potential discrepancies. The analysis provides a summary of Potential Issues, which highlights key metadata problems detected within the document.
Users can click the Analyze with AI button to generate an AI-powered summary that provides an overall description of all identified metadata issues. This summary not only lists each flagged metadata item but also explains the reasoning behind every flag, offering deeper context and clarity.
Users can also filter the list to show either all document metadata or only flagged metadata. Each row contains, the metadata name, description, and and exclude button that allows the user to remove flagged metadata from the analysis.
2. AI Authorship
The AI Authorship Analysis offers detailed insights into the likelihood that parts of a document were generated by artificial intelligence. This feature helps users understand the composition of the document by identifying which sections may have been AI-generated versus human-written.
At the core of the analysis is the AI Authorship Index, a probability-based classification that estimates the presence of AI-generated content. This can be High, Medium, or Low probability.
The analysis also includes key metrics that provide a comprehensive overview. These metrics are the total number of sentences in the document, the number of sentences likely generated by AI (Potential AI Sentences), and the number of sentences most likely written by a human (Human-written Sentences).