RICHLAND, Wash. -- Modern technology has created the opportunity to rapidly access information on a myriad of subjects, often resulting in the generation of hundreds, and sometimes thousands, of pieces of data. However, in many instances, the abundance and variety of information can be overwhelming for decision-makers.
Starlight, an advanced three-dimensional visualization technology, has been developed by Pacific Northwest National Laboratory in Richland, Wash., to help solve the problem of information overload. Already in use by the U.S. intelligence community, Starlight can be applied to a variety of other fields, such as medical data analysis, environmental security and current events monitoring.
Advanced data sources and rapid communication technologies generate massive collections of information rich with facts. However, such an information overload can hide critical details and subtle relationships that are integral to effectively interpreting the data," said Randy Woodson, Pacific Northwest program manager. Starlight provides an easy-to-use interactive tool for exploring and integrating data needed for effective decision making.
For example, an organization like the Centers for Disease Control could use Starlight to help assess potential public health effects that may result from global warming," Woodson said. Starlight could enable CDC analysts to search and interpret massive and diverse amounts of information on global warming trends."
The U.S. Environmental Protection Agency also could use the technology in its comparison of the compliance of Superfund sites to environmental regulations. Currently there are 1,200 designated Superfund sites in the nation and thousands of other potential sites.
The versatility of Starlight enables users to analyze massive volumes of information contained in various media, including structured and unstructured text, maps, digital data, video and even satellite imagery," said John Risch, Pacific Northwest principal researcher. Users can analyze both information content as well as complex interrelations among individual elements of large information collections."
Starlight has two functional components. The first is an information preprocessing and modeling system. This component characterizes information content, modeling associations among input data elements and stores the modeled information for later retrieval and analysis.
The other component is a visualization system/user interface, which is used during exploratory analysis of the information model. The system operates on Windows NT platforms, providing a common and familiar learning and operating environment.
The visualization system graphically represents collections of information as sets of 3D shapes, or glyphs, distributed throughout the 3D work space. Each glyph represents an individual piece of information, such as a document or news-wire report. Topically related items form groupings, or clusters, providing a graphical table of contents to guide general browsing. Analysts can interactively navigate the clusters internally to inspect local conceptual interrelations, or they can view the entire collection to get a sense of general trends in the information.
Analysts also can use different colors and shapes to support the visual cross-referencing of conceptual, or topical, information with other properties of the data. For example, the color- and shape-encoded glyphs can make obvious things, such as similarities or differences in source, author or publication date, among topically related documents.
Starlight represents a new class of information system -- one that has been designed specifically for the visual analysis of complex, multimedia information that is necessary for effective decision-making today," Woodson said.
The U.S. Army Intelligence and Security Command and the Department of Energy are the principal sponsors for the development of Starlight.
For more information on Starlight, contact Randy Woodson at phone: (509) 372-6420 or e-mail: randy.woodson@pnl.gov.
###