An Enriched Meta-Information Schema for Astronomical Databases

S. C. Calvo, N. E. White

NASA Goddard Space Flight Center

T. A. McGlynn

Universities Space Research Association

M. M. Duesterhaus, C. A. Rosen, E. J. Sabol

Hughes Corporation

Abstract:

1. Background

The HEASARC enriched meta-information schema, or metabase, was conceived as part of the effort to migrate the various types of astronomical information held at the HEASARC to a common relational database management system.

There were several objectives in developing the metabase. The first was to allow client-server access to the data and information systems. This would enable the use of existing client interfaces such as World Wide Web browsers. as well as the development of any special-purpose interfaces which might be appropriate.

Another goal was to provide more information in order to facilitate selection and use of data from multiple missions. In a multi-mission archive facility like the HEASARC, it is also important to decouple the information structure from the data file structure to allow easier long-term maintenance.

Finally, the entire system was intended to be portable and to enable exchange of data and information with other institutions.

2. Approach

The schema for the HEASARC database is structured in three layers. The metabase layer contains information about the catalog tables and about the data product layer. The catalog layer contains astronomical catalogs, proposal information, etc. The data product layer contains information identifying individual ``granules'' of data.

The metabase is used by HEASARC information systems to inform the user about available data and information, determine which tables to query in response to a user's request, and format the query and results appropriately. It can also be used in data exchange.

3. Metabase Overview

The metabase consists of seven tables, four (ZZGEN, ZZEXT, ZZPAR and ZZREL) which describe the catalog layer, and three (ZZDP, ZZDPTYPES and ZZDPSETS) which describe the data product layer. The function of the metabase tables are as follows:

• ZZGEN: Describes individual tables in terms of their generic database attributes. This table might be used as is to describe tables in any application domain.

• ZZEXT: Designed to contain discipline-specific information using parameter name/parameter value pairs. For maximum flexibility, this approach avoids hard-coding discipline information into the table schema. Instead, conventions may be established for a given discipline or sub-discipline.

• ZZPAR: Contains information about individual columns in a given table. Again, the attributes have been designed to be generally useful rather than discipline-specific.

• ZZREL: Describes relationships between tables and their parameters. Intended to accommodate joins.

• ZZDP: Relates abstract data product tags to locations given as URLs. This provides the potential of pointing to services, etc., as well as files, and also builds in the ability of access across multiple sites.

• ZZDPTYPES: Intended to accommodate the use of aggregate data types. In the astronomical domain, typical data product types might be: image, light curve, spectrum, etc. An example of an aggregate data product type might be @BATSEFILE, which is a FITS file that might contain both a light curve and an image.

• ZZDPSETS: Defines generic sets of data products, and describes the structure of the data product tags for the set.

4. Development Status

The portion of the metabase describing the catalog-level tables is populated and in use in the current release of the HEASARC W3BROWSE application.

Population of the data products portion of the metabase is underway, and it is expected to be used in the next release of W3BROWSE early in 1996.

Efforts are beginning to test various interoperability concepts, including using the metabase to facilitate responding to queries from outside applications.

5. For More Information

• An introductory software report was published in the May 1995 ``Bulletin of the American Astronomical Society''.

• Detailed documentation on the HEASARC Database System and associated client-server software can be found at:

  http://heasarc.gsfc.nasa.gov/docs/software/dbdocs/dbdocs_top.html

• Please contact Sherri Calvo, sherri.calvo@gsfc.nasa.gov, with any questions or comments.