The Public Record Office hosted another of its immensely useful EAD
Users' Day on January 29th.  

Meg Sweet (PRO) was one of two international representatives on the
working party converting the beta version of EAD DTD to version 1.0.  She
commented on the main changes between the two versions, particularly
the <add> tag, which can now be used at any level of description, and
the <abstract> tag, which has been recommended to be mandatory in
ISAD(G), and which is helpful where the title of an archive is not
necessarily informative.  The aim of the tag is to give context and
content information at the first or second screen to be seen. It
should give a very brief statement of biographical or historical
information and scope and content information.  She also mentioned
what are seen as failures to map with the Dublin Core set of metadata
elements.

Bill Stockting (PRO) went through the various sources of information
available about EAD, and guided  us through the Library of Congress's
site (URL <http://lcweb/loc/gov/ead/>).  He gave a very useful
summary for those beginning EAD.  He asked for comment on a PRO offer
of training in elementary EAD.  He had in mind a two day course to
cover the basics. Please contact him to register your interest, so
that he can estimate the demand.  

Lesley Richmond (University Archives, University of Glasgow) reported
a very interesting meeting she had attended at Yale before Christmas,
which had discussed proposals for an Encoded Archival Authority DTD. 
An Authority Record, which would define an entity, and be consistent
over space and time, would be of great use not just to archivists,
but to museums, libraries, and academics, and possibly commerce. 
Once established the authority record could have links to archives
related to the subject of the record.  She used as her example the
papers of Lord Kelvin, which are in Glasgow and Cambridge.  This is,
of course, an expanded international NRA, where the Register entries,
compiled to agreed standards (ISAAR (CPF) and a DTD), would link to
the electronic finding aids.  Glasgow is contributing to the National
Names Authority Files project and Lesley demonstrated some of the
records they had compiled.  She also showed how they were using the
records to get the correct form for names in their finding aids. 

Paul Watry (University of Liverpool) talked about attribute sets for
Z39.50.  Z39.50 is an information retrieval protocol, which enables
complex search and retrieval from distributed databases and
information resources.  It enables descriptions to be navigated from
collection level to specific files.  The 'protocol' is able to do
this only if everyone conforms to the same rigorous set of rules. It
does this by means of 'attribute sets' or labels for data on which
everyone everyone has agreed.  If the rules are not followed
cross-domain searches won't work.  The Implementation Group for
Z39.50 (ZIG) is currently drafting an architecture or more formal
guidelines for the attribute sets.  Like other participants Paul was
uncertain about mapping to Dublin Core elements.  He stressed the
three separate sets of information needed for Z39.50:  content
(compiled to rules such as MAD, RAD, ISAD(G), and APPM); format (EAD,
MARC-AMC, TEI); and delivery (paper or electronic).  Electronic
delivery had a further format to decide to enable delivery (HTML,
DynaWeb, and Z39.50).  Most archivists working in the British Higher
Education sector are based in libraries which conform to the world
wide standard of MARC and it is vital for these archivists to
establish a precise relationship between MARC-AMC and EAD.  Paul
raised other issues that need to be discussed. For example who would
develop and own an attribute set; who would ensure that guidelines
and best practices for EAD attribute sets are created and
disseminated; how could we ensure co-ordinated development amongst
the different communities (e.g. CIMI, TEI, MARC). 

Julie Allinson (University of Liverpool) spoke about two of
Liverpool's pilot projects for converting existing catalogues to
EAD, from a database (Procite) and from a typed, much annotated
calendar. The calendar of the Cunard papers was over 1,000 pages
long.  In three weeks Julie analysed it and marked up different
levels of description with different marker pens.  A key explained
what each colour meant and what tags to use for it, e.g. pink
indicated series level information. The text plus templates were sent
for off-shore re-keying at a cost of stlg1/page (plus startup fee of
stlg500).  A month later the text was returned, with tags inserted.  The
second project used Procite, a commonly used bibliographical
database.  It had been originally chosen for writing a finding aid
largely because it had been available and was familiar to the
archivists.  Like all databases the information had been inserted in
fields, and this enable Pete Johnston (then at Liverpool; see below)
to match this information with tags and convert it to EAD format. 

Matthew Hillyard (PRO) spoke about TOPCAT, the PRO's project to get
its top-level descriptions converted to EAD.  350,000 pages of lower
level information had already been converted (PROCAT) and was already
available to researchers.  The information in TOPCAT had been taken
from the PRO's printed guides in 1994 and entered into Oracle 6, a
relational database.  It has now been converted to a later version
of Oracle and this has been converted to EAD.  The project is
expected to be launched in April.

Pete Johnston (University of Glasgow) talked
about his experiments with Extensible Markup Language (XML) and
Extensible Stylesheet Language (XSL).  The University Archives is
compiling collection level descriptions.  A Microsoft Word template,
based on ISAD(G), has been compiled, with paragraph styles
corresponding to the ISAD(G) elements, and character styles used to
identify sub-paragraph level EAD elements such as dates and
cross-references.  Word cannot nest styles, but nesting can be
implied through the appropriate naming of styles.  Pete had written a
Word macro which took the styled document and added tags to generate a
complete EAD-encoded document.  Pete stressed the need for information
to be structured.  Once documents have been converted to EAD and are
SGML-compliant they can be converted to XML. The greater simplicity of
XML will end the necessity for plug-ins such as Panorama and for
converting EAD documents to HTML. 

In the lunch break we were able to play with the PRO's TOPCAT, as
well as other EAD sites, as well as chewing over problems with other
EADers.  It was an excellent day, and we are all very grateful to the 
PRO for holding it.  I hope the PRO can continue to
hold days like this.

I will send a copy of this to Heather for the Society of Archivists'
<Newsletter> but I suspect it will have to be considerably
shortened.

CW


Christine Woodland
Acting Archivist
Modern Records Centre, University Library
University of Warwick, Coventry, CV4 7AL, UK
tel. (01203) 524495
fax. (01203) 524211
email  [log in to unmask]
URL http://www.warwick.ac.uk/services/library/mrc/mrc.html


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%