Intelligence Analysis

Intelligence analysts are faced with an ever growing variety, volume, and velocity of data. Studies have shown that roughly 80% of data is unstructured, requiring advanced, scalable text analytics capabilities to make sense of it and apply the gained insight to enhance national security.

Knowledge Discovery – Beyond Search

Traditional search is useful when we know exactly what we are looking for, but what if we have a more general question like “Tell me everything about this new suspect named X.”

NetOwl can analyze texts that mention this suspect X and construct a “biography” automatically thanks to its entity, relationship, and event extraction. Such a biography may contain the person’s relationships to other people and organizations, person’s activities, and person’s physical characteristics. Instead of reading through a large amount of texts, analysts can focus on verification and deep analysis of what NetOwl discovers.

Link Analysis – Connecting the Dots

One of the core functions in many intelligence analysis applications is the discovery and understanding of the “network” an individual or group of interest is a part of. Networks can be based on relationships between two entities, such as familial relationships between two people or membership relationships between a person and an organization. NetOwl’s relationship extraction helps find such relationships in texts.

Networks can be also based on entities participating in events. For example, we may be interested in individuals holding a meeting or an organization shipping an artifact to another company. NetOwl’s event extraction enables discovery of different types of events from texts and establishes relationships among their participants.

Geospatial Analysis – From Texts to Maps

Some analyses are done best geospatially. This is no exception to analysis of texts, but there is one problem – texts do not usually come in a format usable by geospatial analysis tools such as Esri ArcGIS and Google Maps. NetOwl’s Smart Geotagging solves this problem.

By geotagging location entities mentioned in texts and then associating other entities like people, organizations, and events, to those locations through semantic extraction, NetOwl produces the structured output needed to build a rich geospatial index of unstructured data. In its simplest form, an analyst can now select an area on a map and ask for texts that are relevant to this area. In a more advanced case, analysts can ask for events of interest displayed on a map so as to discover trends or anomalies.