Search

A major goal of the Virtual Observatory is to build tools for exploring and extracting new information from heterogeneous (and possibly very large) data sets. There are two different facilities that a proper data mining facility tool should provide: first, a selection tool to locate relevant datasets; second, a tool to combine information from the selected datasets. We present data mining tools that have been developed at CDS, as practical implementations of a V.O. application (the first development was begun in the frame of the ESO/CDS data mining project). These tools rely extensively on long-term standards and emerging metadata (e.g. UCDs). They include: a selection tool to perform catalogue selection in the VizieR data federation on various criteria (keywords, UCDs); a positional cross-correlator to perform matching between different datasets. [ABSTRACT FROM AUTHOR]

1. Introduction2. UCD Exploration3. Catalogue Selection4. Automated Column Comparisons5. Other Applications6. Conclusions [ABSTRACT FROM AUTHOR]