DTDs Together: TEI, CIMI & EAD

[Amy Tracy Wells]

Please excuse duplication.  The following are my notes from a CNI
(Coalition for Networked Information) project briefing I attended on
December 6th, 1996. I hope you find them useful. All errors are mine.  
Amy

Amy Tracy Wells, Internet Librarian	awel at cs.wisc.edu
InterNIC Net Scout			Voice (608) 263-2611 
University of Wisconsin-Madison		Fax (608) 262-9777
Computer Science Department
1210 W. Dayton St.
Madison, WI 53706
http://www.cs.wisc.edu/scout -- Net Scout Services Home Page
***

DTDs Together: TEI, CIMI & EAD

The following participated in a CNI Project Briefing, chaired by David
Green, Executive Director, National Initiative for a Networked Cultural
Heritage (NINCH).

Allen Renear, Director, Scholarly Technology Group, Brown University

John Perkins, Project Director, Consortium for Computer Interchange of
Museum Information

Daniel Pitti, Librarian for Advanced Technologies Projects, University of
California, Berkeley

Standard Generalized Markup Language, mooted as early as the 1960s but not
adopted as an international standard until 1986, is the coding or tagging
of structural information within a document or image. The development of
SGML and subsequently, DTDs, Document Type Definitions, arose out of a
need for inter-operability and search capabilities across various
collections and systems. 

The importance of this session was that the three separate but related
critical DTDs used in the cultural community were presented together,
displaying differences and similarities, as a first step to building
harmonization between them.

Three Document Type Definitions, DTDs, were reviewed.
        1) Text Encoding Initiative or TEI
        2) Consortium for the Computer Interchange of Museum Information
                or CIMI (pronounced "see me")
        3) Encoded Archival Description or EAD

Allen Renear discussed the Text Encoding Initiative or TEI. "TEI" is the
name of a project and a set of guidelines. It has been designed as an
interdisciplinary and collaborative effort to "allow scholars to identify
what they want to identify" (Renear) within various sources. TEI supports
conformance levels within its tripartite structure. The latest set of
guidelines, P3, was published in 1994.
        Core Tag Set
                1) core
                2) header
        Base Tag Sets
                1) prose
                2) verse
                3) drama
                4) spoken
                5) dictionaries
                6) terminology
                7) general
                8) mixed
        Additional Tag Sets
                1) linking
                2) analysis
                3) feature structures
                4) certainty
                5) transcription
                6) text criticism
                7) names.dates
                8) network
                9) figures
                10) corpus
(See www.uic.edu/orgs/tei for more information)

John Perkins discussed the Consortium for the Computer Interchange of
Museum Information, CIMI, which began in 1980 to address the needs of the
museum world. Currently, a project demo, CHIO, or the Cultural Heritage
Information Online exists to "ensure long-term interchange of museum
standards" (Perkins). The purpose of the project is two fold 1) search and
retrieve information and 2) provide integrated access to information i.e. 
libraries, museums, art galleries, etc. across organizations and type of
organizations. 

SGML was chosen as a method to make documents "databaseable", for
varieties of output, for navigation of complex documents and to capture
content at high levels of richness. The design principles are to render
source document, build on previous standards, to provide online access to
relevant parts of a document (as opposed to the document in its entirety)
and to distribute information resources. CIMI incorporates the basic tag
sets but extends the Base tags.
(See ftp.cimi.org/pub/cimi/CIMI_SGML/cimidtd.ps
and/or www.cimi.org/cimi for more information)

Daniel Pitti discussed the Encoded Archival Description, EAD, which began
in 1993 to study the feasibility of developing an encoding standard for
archival finding aids . The Library of Congress is the maintenance
agency for this standard.

EAD seeks to address a wide range of institutions and implicitly document
types) from "libraries, archives and museums to provide intellectual
control of and access to archival collections of primary source materials" 
(Pitti). The primary characteristic is "progressive hierarchical
description frequently involving many lists" (Pitti). EAD is being used by
a wide range of information specialists and "increasingly...the public".

(See ftp.loc.gov/pub/ead
and/or ftp://library.berkeley.edu/pub/sgml/ead for more information)

All three enjoy international support.