Data Mining Techniques Applied to the GNAT Library Archive (Poster abstract)

Volume 37 number 2 (2009)

Erin M. Craine

Abstract

(Abstract only) The Global Network of Astronomical Telescopes (GNAT) is in the process of developing a large library of data for its newly discovered variable stars. This revised archive has been recently opened to the public, is largely unexamined, and provides a valuable resource for data mining. The GNAT archive is constantly growing as survey imaging continues; in 2009 the survey imagery is estimated to contain in excess of 150,000 new variable star entries selected from more than about nine million observed stars. The algorithm used for identifying variable stars in these images yields examples of nearly all known classes of such stars. One of the key goals of developing data mining techniques is to be able to narrow down this large volume of data to specific stars of interest. We discuss the nature and content of the GNAT variable star archive, including a discussion of limitations and boundary conditions that affect the data. We present some basic ideas of data mining, examine specific approaches to the GNAT archive, and provide examples of how to extract information related to these variable stars.