KModelC
Investigators, analysts and cybersecurity teams reviewing AI-generated investigative information
Category
Investigation & analytics
Owner partner
HUA
Related use cases
UC1.1, UC1.2
Tool summary
How can investigators connect information hidden across images, audio and text? KModelC extracts relevant facts and relationships into a shared structure, helping analysts explore potential links in investment fraud and other financial investigations.
The challenge
Investigators work with information scattered across images, audio recordings and written material. Identifying people, objects, events and relationships across these sources takes time, while important connections can be missed. They need a way to bring relevant details into a consistent structure that can support analysis across different types of investigation.
What it does
KModelC processes images, audio and text to extract relevant entities and relationships. It organises the extracted information according to a shared ontology and represents it in RDF format, so that findings from different sources can be connected and examined together. This structured output can help investigators explore links that would be harder to identify across separate files. The approach can support different investigation scenarios, including online investment fraud and the analysis of suspicious financial activity.