In this short How To video we are going to explore the Text Analytics capabilities built into ADF’s digital forensic software with the integration of Rosoka. ADF is bringing the power of Rosoka for automated entity extraction and language identification with gisting, through a tightly integrated user experience in the Rosoka Add-on.
The Rosoka Add-on runs locally on the investigator’s computer, processing documents in over 200 languages to identify entities, and locations in unstructured documents.
The Rosoka Add-on allows front-line operators to speed their intelligence gathering with the ability to view translated data from structured and unstructured evidence so they can make better-informed decisions starting on-scene.
Rosoka is built into Triage-G2, Triage-G2 PRO and available as the Rosoka Add-on for Mobile Device Investigator, Digital Evidence Investigator, and Triage-Investigator or our bundled PRO tools.
After completing a scan the Classifier will run Entity Extraction automatically. This process can be paused and resumed if necessary. Entity extraction will run after image and video classification.
Once completed the entities can be filtered in the table where entities were located and also in the entities tab on a specific file. When looking at a specific file with entities the “gloss” or “gist” is shown as well as the original file.
Along with documents, Entities can also be located in Email and Messages.
The Rosoka Add-on identifies
- Over 3 dozen different entity types
- Over 500 different relationship types
- Truly multilingual platform
- Simultaneous extraction in over 200 different languages
Simply put, gisting is the use of Natural Language Processing (NLP) to translate foreign text into English so a forward operator can get an understanding of the original content’s meaning. To do this, the translation doesn’t need to be perfect. It just needs to translate the general meaning of the foreign text so that operators get the "gist".
The Rosoka Add-on is a valuable tool when needing to find the main point of a large document in a time-sensitive matter.