You have to first understand that there is a plethora of information locked away in unstructured sources. The ability to get to that data is somewhat convoluted.
Saffron’s experience with national security pre-9/11 has highlighted this ‘big data’ problem particularly with data coming from news, email and many other sources of unstructured content. Saffron figured out that you could use plenty of tools, called ‘entity extractors’, as one element of dealing with text. So if you read documents and comments it is not just about storing the text or data, but about reading it in some sense of finding what:
What things are?
What are the parts?
What are the problem descriptions?
Who are the people?
What are the companies mentioned?
What products are mentioned?
When you start using these ‘entity extractors’ to find, and mark, the people, places and things the situations, problems, actions and outcomes give Saffron a memory base. This leads to opportunities to learn about those things and connections between them, the connections between people and the connections between problems and solutions.
It’s in the text and part of the solution to address text is first to use these technologies of marking people, places and things, situations, actions and outcomes. If you mark these then Saffron as a memory base can help you learn about all those people, places and things, situations, actions and outcomes and how they are connected. That’s the knowledge, and the connectivity of the data. It is not just the raw data itself and this is how Saffron gets into the unstructured text to exploit it.