- November 30, 2018
- Posted by: Alexis Low
- Category: Articles
IAI will continue development of a system to auto-extract data relevant to significant events and to archive patterns. A key innovation of the Scalable Event Extractor for Multi-Level Event Data and Pattern Archiving system (SEE) is a large class of analytics that extracts valuable and important information from raw data. Another innovation is a two-level event data and pattern archiving architecture using heterogeneous information networks (HINs) and EventCube. This enables fast information retrieval, efficient information storage, and powerful analytics for question and answers.
SEE’s intelligent algorithms can summarize the large amounts of data collected by Naval Intelligence Surveillance and Reconnaissance (ISR) systems and store it efficiently so that the data does not need to be discarded after some time. The system will automatically capture and retain only valuable and important information by archiving it in a condensed form that preserves information on entities, relationships and context from the raw data. SEE’s intelligent algorithms build event context awareness and preserve essential data in a form for efficient recall and analytics. Running multi-INT multi-source analytics efficiently will allow commanders and analysts to work with many years’ worth of data.
SEE will have a mature, intuitive and responsive human computer interface. SEE will be deployed as a decision aid toolkit to significantly improve the Navy and Marine Corps’ capability to create actionable intelligence from billions of documents, nodes, and edges. Internet search engines would also benefit from the maturation of data retrieval based on embedded space showing relationships of content and the second or higher order association (degrees of separation) retrieval techniques being developed in SEE.
For more information on SEE or to find out how IAI can develop a custom auto data-extraction solution for you, contact us at firstname.lastname@example.org.