4 Urania

The idea that the internet could be used to link sources of astronomical information into a unified environment is more than a decade old; it was fully expressed in the planning for the old ADS (Squibb & Cheung 1988) and ESIS (Adorf et al. 1988) projects. These early attempts were highly data oriented, their initial goals were the interoperability of different distributed data archives, primarily of space mission data.

Astronomical data is highly heterogeneous and complex; essentially every instrument has its quirks, and these must be known and dealt with to reduce and analyze the data. This quirky nature of our data essentially prevented the establishment of standardized tools for data access across data archives.

The new, hyperlink connected network data system for astronomy is based on the highest level of data abstraction, object names and bibliographic articles, rather than the lowest, the actual observed data in archives. This change in the level of abstraction has permitted the creation of a system of extraordinary power. This new system, still unique amongst the sciences, has been dubbed Urania (Boyce 1996), for the muse of astronomy.

Conceptually the core of Urania is a distributed cross-indexed list which maintains a concordance of data available at different sites. The ADS maintains a list of sites which provide data organized on an article basis for every bibliographic entry in the ADS database. The CDS maintains a list of articles and positions on the sky for every object in the SIMBAD database. The CDS also provides a name to object resolver. The possibility for synergy in combining these two data systems is obvious; they have functioned jointly since 1993.

Surrounding this core, and tightly integrated with it, are many of the most important data resources in astronomy, including the ADS Abstract Service, SIMBAD, the fully electronic journals (currently ApJL, ApJ, ApJS, A&A, A&AS, AJ, PASP, MNRAS, New Astronomy, Nature, and Science), NED, CDS-Vizier, Goddard-ADC, and the ADS Article Service. All these groups actively exchange information with the Urania core, they point their users to it via hyperlinks, and they are pointed to by it.

The astronomy journals which are not yet fully electronic, in that they do not support hyperlinked access to the Urania core, also interact with the system. Typically they provide access to page images of the journal, either through PDF files, or bitmaps from the ADS Article Service, or both. Bibliographic information is routinely supplied to the ADS, and the SIMBAD librarians routinely include the articles (along with those of the electronic journals) in the SIMBAD object-article concordance.

While most data archives are not closely connected to the Urania system there are some exceptions. For example the National Center for Supercomputing Application's Astronomy Digital Image Library (Plante et al. 1996) connects with the ADS bibliographical data via links which are papers written about the data in the archive. SIMBAD connects with the High Energy Astrophysics Science Archive Research Center (HEASARC) (White 1992) archive using the position of an object as a search key, HEASARC has an interface which permits several archives to be simultaneously queried (), and a new data mining initiative between CDS and the European Southern Observatory (ESO) (Ortiz et al. 1999) will connect the Vizier tables with the ESO archives. Several archives use the SIMBAD (and in some cases NED) name resolver to permit users to use object name as a proxy for position on the sky, the Space Telescope Science Institute (STScI) Digital Sky Survey (Postman 1996) would be an example. The Space Telescope-European Coordinating Facility archive (Murtagh 1995) allows ADS queries using the observing proposals as natural language queries, and the Principal Investigator names as authors.

The establishment and maintenance of the Urania core represents a substantial fraction of the ADS service. SEARCH discusses the user interface to the set of hyperlinks, ARCHITECTURE discusses the methods and procedures we use to implement and maintain the links, and DATA discusses the data sharing arrangements we have with other groups, and presents a complete listing of all our data sources.