Glossary
Every classifier description has the same groupings with various points below: Categorization and Details.
Categorization
- Category: The Category in the Patrol UI.
- Sub-Category: The Sub-Category in the Patrol UI.
- Type: The general type of analysis performed by the classifier. Types includes Text, Link, Image, Video, and Account Analysis.
Details
- Methodology: This is the general category for how the classifier flags contents. More information can be found in the Technical section.
- Directionality: This is the direction of the content that the classifier will scan.
- Entities: These are the Patrol entities for which users can enable the classifiers. Entities include Account, Brand, People, and Location.
- Data Source(s): These are the data sources that are fed into the classifier. Sources include LinkedIn, Twitter, Facebook, Instagram, YouTube, and DarkNet.
- Data Filter(s): This is an optional detail that contains any filters to the data sources listed above.
- UI Element(s): This is an optional detail that can contain "Highlight" (the match phrase is highlighted in the UI) or "Mask" (the match phrase is masked in the UI).
- Match Phrase: This is an optional detail that is either "Yes" or "No" depending on whether the classifier returns a match phrase for flagged content. Our roadmap has all classifiers returning a Match Phrase in the near future.