Metadata Processing in Buckets
				
		    Steve Robbins <s.w.robbins@larc.nasa.gov>


The buckets element inn NCSTRL+ [1] provides a method to obtain
metadata for the bucket in various metadata formats. This capability is
currently not implemented.

In my project, I will begin to implement this capability. One interesting
aspect of my project will be where in buckets to place this capability. A
number of approaches are possible: (1) Place the metadata conversion
capability in the bucket itself, (2) let the bucket communicate with a
network service to obtain the conversion capability, and (3) preconfigure
and store all of the supported metadata forms in the bucket when it is
created. A hybrid approach combining all or some of these or other
approaches may also be possible.

This capability will probably be implemented as a Perl 5 module. I will
investigate other metadata projects such as [2] and [3] and reuse any
applicable components.  The following metadata formats will be
implemented as time permits: RFC-1807, Dublin Core, Refer, and Bibtex.
The implementation should allow additional formats to be added in the
future.

1.  M. L. Nelson, K. Maly, S. N. T. Shen and M. Zubair, "Buckets: Aggregative,
    Intelligent Agents for Publishing," Old Dominion University CS TR-97-41,
    December 1997.

2.  D. Jacobsen, "bp, a Perl Bibliography Package",
    http://www.ecst.csuchico.edu/~jacobsd/bib/bp/index.html

3.  A. Paepcke, "InterBib: Bibliography-Related Services",
    http://Walrus.Stanford.EDU/~testbed/interbib/