InferX technology, originally developed for the military, enables scoring of suspicious behavior for investigative data mining and counterintelligence. Our pattern recognition capabilities enable targeting under uncertainty via collective inferencing — which allows for seemingly meaningless data to become meaningful by drawing inferences from large data sets simultaneously without reliance on guilt-by-association algorithms — which tend to implicate innocent citizens.
Our technology focuses on the analysis of shadowy and unusual interactions and behavior in order to generate risk scores from transactions, conditions, events and sequences — as well as drawing out deviations from normal behavior. This is done in a totally anonymous manner ensuring privacy to innocent citizens. In addition, only suspicious behavior is targeted, meaning only a minuscule of records are pulled for elevated inspections — whether it's containers, events or people.
Analytical risk networks can be created using our technology for the identification of suspicious patterns of behavior for self-adaptive counterintelligence systems. Models and profiles can be created from multiple data sources in a totally anonymous manner without the need to centralize or move any data from departments or agencies.
InferX technology improves detection of threats and events where an embedded entity is not available for social networking analysis using guilt-by-association algorithms for the discovery of 'who knows whom where and when' matching. InferX modeling algorithms are different in that they can be used to discover suspicious patterns, outlier or anomaly behavior as well as for the extraction of key concepts from any kind of textual data: documents, emails, chat logs, etc.