How can we analyze large batches of disparate metadata effectively and efficiently? How can we use visualization tools to create meaningful presentations of metadata for resource discovery and use? Dushay will present a number of explorations, relating them to the traditional library context, to NSDL objectives and to broader metadata analysis and resource discovery objectives.

Naomi Dushay is a research programmer/analyst for the National Science Digital Library. She has been doing digital library research since 1997, contributing to projects such as NCSTRL, Fedora, and PRISM (DLI-2). She was a member of the CU library department currently known as DLIT from 1991-1997.

Naomi Dushay described recent experiences in dealing with large quantities of metadata coming into the NSDL. As an aggregator of metadata, NSDL faces a wide diversity in the range and quality of metadata collected. With the goal of normalizing this metadata, NSDL has attempted a variety of approaches: sampling; sorting and viewing tools (such as spreadsheets); and more recently, visual graphic analysis tools. This talk focused on the latter technique.

Naomi gave a demo of Spotfire, showing it's usefulness at detecting metadata inconsistencies and potential errors and how it has aided the data normalization process at NSDL.

Naomi's experience with Spotfire and other visual graphic tools has led her to explore the use of such techniques to improve the NSDL front end for users. She described various experiments to enhance information discovery and online browsing using data visualization tools such as Spotfire, Treemap, and Star Tree.