All data-intensive endeavors share the same problem: Our assets, which include data, information, and knowledge, as well as analytics, tools, and applications, are fractured in physically and semantically disparate systems. In short, our data enterprise is broken. As a result, insight is obscured and operations are impeded by our inability to fully search, explore, enrich and exploit those assets. In order to understand the full picture, we must break the data barriers. In this presentation, Dr. Yoakum-Stover describes a practical, Ultra-Large Scale systems solution for unified data storage and processing - one that addresses both the scale and diversity of data and processing without imposing constraints on what data must be or how they should be used. She describes how the core innovation underlying the approach - the Artifact, Data, and Model Description Frameworks (AD&M DFs), can accommodate all data, information, and knowledge regardless of modality, structure, and semantics without data loss or distortion, and in particular how geospatial and temporal information are handled. Her talk concludes with a discussion of an instance of the complete system, the Data and Processing Syndicate, built on cloud compute and storage technology (HDFS, Map Reduce, Accumulo) that her engineering team is currently building for the US National Geospatial-Intelligence Agency.
|
|
Susanne Yoakum-Stover
Executive Director, Institute For Modern Intelligence
Suzanne Yoakum-Stover is a data scientist dedicated to developing Modern Intelligence - the science, practice, and governance of intelligence at Ultra-Large Scale (ULS). Since earning her Ph.D. in physics from Stony Brook University in 1992, she has contributed, both as a research scientist and as an educator, to a range of technical fields including physics, artificial intelligence, medical imagi [read more]
|
|
|