Decision makers are regularly overwhelmed by the volume, veracity, and variety of multi-source data. In addition, machine learning (ML) systems can be impacted by collection bias and typically produce uncertain results, even when collection bias is not an issue. To address these data volume, variety, and uncertainty issues, decision makers need a platform that combines uncertain data, uncertain ML predictions, and contextual data to form actionable situation knowledge that they can employ to make proactive, timely decisions.
Seer provides a scalable, web-based solution through which noisy, incomplete, multi-source data can be fused to support real-time situation awareness and predictive analysis. This state-of-the-art, multi-hypothesis, context-ware abductive reasoning framework identifies and exploits patterns hidden in complex data. Through pattern of life (POL) analysis, context propagation, and more, Seer minimizes false alarms while predicting events and situations. Additionally, Seer includes a sophisticated value of information (VoI) calculus to identify missing evidence and quantify its information gain towards resolving situation hypotheses.
For more information, download our datasheet.
Semantic Seer is a web application that mines semantic graph databases for relevant indicators and activity patterns. It automates the discovery of relevant indicators and situations to support predictive analysis and can also support causal analysis to assess the root cause of anomalous events. In addition to supporting a much richer event description language and VoI calculus, Semantic Seer has been a research platform for a variety of DARPA projects, to include situation model learning, knowledge and pattern discovery, and semantic-based fuzzy query.
For more information on Semantic Seer, download our datasheet.
Through situation modeling, Seer aids users in better understanding and visualizing their data by letting them create fuzzy constraints of anticipated events within the system and combining multiple events to form a Situation model. Seer monitors the model and generates situation hypotheses and alerts, providing a predictive analysis of the situation as it unfolds. A full suite of development tools helps users understand and visualize data as well as explore and validate situation models. Model lifecycle management is simplified within Seer, and the platform allows users to create, review, approve, revise, and retire models.
For more information on predictive analysis and situation modeling capabilities, download our datasheet.
Relevant information can come at any time, so ensuring that the right people and processes are notified as soon as possible is critical to success. Seer supports real-time truth maintenance on all situation hypotheses and allows users to associate custom business rules and logic with situation models. These rules trigger custom actions such as alerting users of important discoveries via emails or text messages, intersystem notifications, requests for missing evidence, and much more.
For more information on real-time alerts, download our datasheet.