NET NEWS: Linking a Universe of Databases

Science  27 Aug 1999:
Vol. 285, Issue 5432, pp. 1323c
DOI: 10.1126/science.285.5432.1323c

Imagine a search engine that can dig into a slew of scientific databases scattered across the Internet, then return the info as downloadable data. Such a tool is in the works, for instance, to troll the specimen collections of museums (Science, 7 May, p. 888). And Robert McGrath's group at the National Center for Supercomputing Applications (NCSA) in Urbana-Champaign, Illinois, is developing software that can do the same for astronomy databases.

So far, the NCSA team's search tool, known as Emerge, is only hooked up to a few databases—including NASA's astrophysics abstracts collection and the NCSA's own Astronomy Digital Image Library. Using the Emerge software, astronomers can search these resources not only by title or topic, but also by the position of a celestial object—say, a comet or galaxy. Eventually, the search options will include properties such as brightness and wavelength, says McGrath, who discussed the project last week at the Digital Libraries '99 meeting in Berkeley, California. The data come back in a series of files, like Web pages but in a special astronomy markup language, that includes a small image and a description of the data. The user can then download and work with the relevant portions.

Before the team can add more databases, a group of astronomers at NCSA, NASA, and elsewhere needs to hash out a standardized format for describing, say, a table or an image. “A lot of the problems tend to be sociological rather than technological,” says Robert Wilensky, director of the digital libraries project at the University of California, Berkeley. McGrath's group is also adapting the same tool for the National Cancer Institute, which wants to link more than 50 databases with such information as case histories of genetic disorders.

Navigate This Article