Decision makers are acutely aware that they are often swimming in sensors and drowning in data. Complicating this is the fact that the data is often incomplete, inaccurate, or simply missing. As a result, much of this data simply “falls to the floor”, never to be seen or analyzed by those for whom the data was collected. What is needed is a capability to turn this data into actionable information that can be used to make proactive, timely decisions. To address this critical need, ISS has developed Seer, a fuzzy Complex Event Processing (CEP) system that can successfully process noisy, incomplete, multi-source/multi-INT data in support of near real-time decision making.
Seer is a “fuzzy” CEP solution designed to support decision makers by identifying and exploiting patterns hidden in complex data. Seer can operate in forensic mode against historical data, in near real-time mode for proactive decision making or in combination. Seer makes use of advanced fuzzy information fusion algorithms to successfully exploit observation data that may only partially satisfy an event description in time, space or other relevant dimension.
Through the use of sophisticated context propagation, Bayesian reasoning, and spatiotemporal analysis, Seer is able to provide both predictive awareness of upcoming events and likelihood analysis for events that may have occurred but were not evident in the collected data, all while minimizing false detections. Custom business rules and logic can be associated with Seer models and executed against the data matched against a model’s event descriptions.
Relevant data can come at any time and it’s important to know about it right away. Seer recognizes the importance of communication in multiple ways. Customizable alerting allows the end user to be notified via email, text messages, popups, WebTAS displays, JMS messages and more.
Semantic Seer mines graph databases and Resource Description Framework (RDF) triple stores for relevant indicators and activity patterns. Semantic Seer contains a sophisticated machine-learning capability for automating the discovery of relevant indicators and activity models. In addition to predictive analysis, Semantic Seer supports causal analysis to assess the root cause of an anomalous event.
Model development allows a user to create constraints of an anticipated event; a series of these events form a model. Seer monitors the model and generates alerts based on the information, providing a predictive analysis of the situation as it unfolds. A full suite of model development tools helps users understand and visualize data, explore and validate models. The model lifecycle management tool includes model creation, review, approval, revision and retirement.