AI7-Minutes of the 3rd TEI AI7 meeting of 1991-11-15/16
 
In attendance:                               Observers:
Sue Ellen Wright                             M. Leamen, ISO/CS
Gregory Shreve                               G. Herzog, DIN
Gerhard Budin                                K.-D. Schmitz, Univ. of Saarbrücken
Absent:                                      H. Hjulstad, RTT
Alan Melby
Dick Strehlow
 
Ms. Wright presided in the absence of Mr. Melby.  She began the meeting with a
short introduction to the work of TEI AI7 for the benefit of the observers
present.
 
1.      Revision of work to date
 
The members of AI7 discussed changes made in TEI-TERM since the last meeting
in Oak Ridge, USA, most of which are contained in the current version of
papers delivered by Wright and Melby at the Conference on Terminology and
Documentation in Hull, Canada, and at ASLIB in the UK.  (AI7 W14 and AI7 W20,
respectively)
 
2.      ISO activities
 
Mr. Leaman circulated a DTD developed by ISO/IEC/JTC 1/SC 18/WG 8 for handling
terminological data for the publication of standards (Attachment # 1).  He
explicitly welcomed the activities of TEI AI7 as a valuable contribution to a
broad ISO strategy for developing a whole range of SGML support facilities.
 
3.      TEI Guidelines (P2)
 
Since AI7 has to present a chapter for the final version of the TEI Guidelines
(TEI P2) very soon, discussion began on what to include in this chapter,
centering on the formal requirements provided by the TEI editors.  Mr. Budin
stated that all documents created so far by the group should be condensed and
merged into a new document specifically designed for the TEI Guidelines.  This
work had to begin immediately because the final document was due in Chicago in
electronic form by 28 November 1991.
 
4.      Future work
 
After completion of the P2 document, empirical tests must follow.  Operators
of large Terminological DataBases (TDBs) and developers of Terminology
Management Systems (TMSs) should be encouraged to collaborate with AI7 in
testing the format by compiling conversion routines to export data from their
systems to TEI-TERM format.
 
5.      Current concerns
 
5.1     The group discussed a newly introduced element, the tag <ofig>, the
other form information group.  This tag has been introduced in order to avoid
mixed content models following the <otherForm> tag.
 
Example:
<tig>
    <term> PCDATA </term>
    <descrip> ... <citnRef ...>  </descrip>
    <ofig>
        <otherForm> PCDATA </otherForm>
        <descrip> ... <citnRef ...> </descrip>
    </ofig>
    <xref type=... target=...>
    <date> 1991-11-17 </date>
</tig>
 
5.2     Mr. Hjulstad pointed out the problems inherent in relegating the grammar
data element to the level of an attribute value in the form of specific
grammatical items of information (pos, gen, etc.).  This problem is
exacerbated if individual term elements must be documented for multiword
terms.  As a result of these discussions, the group decided to include a
<gram> tag in the list of tags, in which case the specific types of
grammatical information would become values of the attribute type: Ex. <gram
type=pos> n </gram>.
 
6   Preparation for ISO/TC 37/SC 3 meeting
 
AI7 had submitted several documents to the International Standards
Organization Technical Committee 37 Sub-Committee 3 for Computational Aids to
Terminology (ISO/TC 37/SC 3) in conjunction with a proposal for a future work
item.  The group discussed the strategy to be adopted at the next ISO/TC 37 SC
3 meeting on 1991-11-18/19 in Vienna.
 
Several options were presented with respect to SC3's possible treatment of an
SGML interchange project vis à vis the existing MATER standard (ISO 6156-1987,
"Magnetic tape exchange format for terminological/lexicographical records
(MATER)."
*   submit an SGML format as a Part II to MATER
*   replace MATER with an SGML format
*   create a separate standard parallel to MATER
*   submit an AI7 document to ISO/IEC/JTC 1/SC18 WG 8 as a contribution to be
    included in a Technical Report on SGML support facilities.
 
The group discussed rationales for these options.  Leaman pointed out that
this effort is fully in line with a recent ISO/STACO recommendation for the
harmonization of structures in terminology work and is thus supported by the
ISO Council.
 
The group discussed revisions on the New Work Item Proposal submitted by the
SC 3 Secretariat prior to the SC 3 meeting.  See Attachment # 4 for the final
revised version of the proposal.
 
7.  Practical demonstration of conversion routines
 
Ms. Wright demonstrated two conversion utilities, MMUTS and NORM.  The current
version of MMUTS (MicroMATER UTilitieS) converts a canonical MicroMATER file
to either a flat or a nested TEI-TERM document.  The NORM utility converts a
flat TEI-TERM document to a nested document prepared for interchange.  For
interchange purposes, systems developers need only create a conversion utility
that will transform their native files into a flat format, which will be very
similar to the structure of their existing terminological entries.
Theoretically, the NORM utility would then be able to convert the flat entry
to the interchange format.  Printouts from the various conversion levels are
included in Attachment # 5.  It is important to note that the documents
included in this Attachment were genuinely generated using the respective
utilities, even though the final product of both utilities is identical.
 
8.  P2
 
At subsequent meetings following the official meetings and after the arrival
of Mr. Melby from the Myrdal meeting, the AI7 group prepared the basis for its
contribution to the P2 document.  It was agreed that this document should be a
concise presentation of the basic elements of TEI-TERM and that detailed
implementations of various features and formats should be dealt with in the
tutorial.  These documents accompany these minutes as AI7 P18 (P2) and AI7 W
19.
 
9.  Future meetings and activities
 
AI7 must address the following projects in the near future:
*   Enlist the cooperation of TDB and TMS developers and users in order to
    test the TEI-TERM format for a wide range of applications.
*   Write the tutorial to be included with the final P3 document.
 
Mr. Melby also noted that a recommendation had been made to add another
European to the WG.  The group was of the opinion that the addition of another
qualified European was highly desirable, but that the exclusion at this point
of any of the Americans already active would be impossible.  Group consensus
would suggest that some sort of compromise be reached with regard to funding,
bearing in mind that AI7 is a highly productive contributor to the TEI effort.
 
The status of future meetings will hinge on discussions with TEI management
and on decisions concerning future funding of TEI activities.
 
Respectfully submitted
 
G. Budin and S. Wright
 
 
Coda:  ISO/TC 37/SC 3  Meetings of 1991-11-18/19
 
The following resolutions passed by SC 3 are pertinent to the work of AI7:
 
Resolutions adopted by ISO/TC 37/SC 3, 1991-11-19:
 
B : The member bodies present at the 8th meeting of ISO/TC 37/SC 3 favor the
establishment of an external liaison, category A, to the Text Encoding
Initiative - Analysis and Interpretation Working Group  - TEI/AI7
"Terminological Data".
    Wright will initiate a formal request on behalf of TEI-AI7 to the ISO
    Central Secretariat for the establishment of this liaison.
 
C : ISO/TC 37/SC 3 establishes a working group (WG 1) "Data elements" within
the scope of SC 3. Ms Wright has been appointed as the convener of the working
group.
    This activity is consonant with the AI7 position that AI7 does not itself
    have a formal mandate to standardize a data element directory in the form
    of a list of attribute values.  Wright will pick up on work already
    performed within SC 3.  The questions of combinability and repeatability,
    formerly a component of SC 3 work on data element names, will be subsumed
    in the more formalistic work of WG 3 on the SGML interchange format.
 
D : ISO/TC 37/SC 3 establishes a working group (WG 3) "SGML terminology
applications" within the scope of SC 3. Mr Budin has been appointed as the
convener of the working group.
    This resolution gives rise to the circulation of the New Work Item
    Proposal discussed above as Attachment # 4.  The version provided with
    this document represents the final edited version from the SC 3 meeting.
 
F : The member bodies present at the 8th meeting of ISO/TC 37/SC 3 favor the
establishment of an internal liaison between ISO/IEC JTC 1/SC 18/WG 8 "Text
description and processing languages" and TC 37/SC 3/WG 3.
    The SC 3 Secretariat will initiate this request for liaison.
 
G : The member bodies present at the 8th meeting of ISO/TC 37/SC 3 favor the
confirmation of ISO 6156, "Magnetic tape exchange format for
terminological/lexicographical records (MATER)" during the ballot for periodic
revision scheduled for 1992.
    The effect of this resolution on the future work AI7 and TC 37/SC 3/WG 3
    is multifold:
    *   No immediate changes will be made in 6156.
    *   The data elements list annexed to 6156 will remain valid while a new
        data elements directory is produced.
    *   SC 3 can initiate a revision of 6156 at a later date in coordination
        with SC 3/WG 3 depending on future decisions.  Any future version of
        6156 is likely to modify the magnetic tape stipulation of the original
        standard.
 
Respectfully submitted,
 
S. Wright
 
Attachments:
 
Attachment # 1:        Revised agenda, TEI-AI7 A15
Attachment # 2:        AI7 Documents List, Printed Version of 1991-11-25
Attachment # 3:        ISO DTD
Attachment # 4:        ISO/TC 37/SC 3 N 89 en Rev 2 of 1991-11-19
Attachment # 5:        MMUTS and NORM output

 
*************************************************************************
Attachment # 1 to TEI M18
 
TEI A15: Revised Agenda of 1991-11-15/16
 
Place: NORM, 1st floor
 
Friday     1991-15-91
 
9:00 - 12:30
 
1.  General introduction of TEI members and guests                                            All
    Determine who wants to go on the excursion on Saturday
 
2.  Brief introduction using some of the slides from the Hull presentation                    SEW
 
3.  Recap of developments since the Oak Ridge meeting based on AI7 WG 14
TEI
 
10:30   Coffee break
 
4.  Review the new DTDs
 
12:30   Lunch
 
2:00 - 5:30
 
    Continuation of morning session, if necessary
 
5.  Strategy planning for the TC 37/SC 3 meeting                                              All
 
3:45    Break
 
5:00    Demonstration of MMUTS and NORM
 
Saturday 1991-11-16
 
8:30
 
1.  General discussion to deal with questions generated from the observers, continued SC 3
    strategy discussions                                                                      All
 
11:00
 
2.  Excursion
*************************************************************************
Attachment # 2, AI7 M17, 1991-11-25
 
TEI AI7 Documents
 
AI7 P1     Objectives and Deadlines: Terminological Data
 
AI7 W2     MicroMATER Tagset (NISKO paper)
 
AI7 W3     A Proposed Format for Exchanging Term Files (ASTM Paper)
 
AI7 A4     Agenda of the Clevelnad Meeting 11-13 June 91
 
AI7 W5     Cleveland Minutes
 
AI7 W6     Sample Term Entries
 
AI7 W7     Draft DTD of 1991-07-27
 
AI7 X8     french.wsd
 
AI7 W9     ISO 5426-1983 Code table
 
AI7 W10    Entities or Attributes for Part of Speech and for Structural
           Analysis of Statements of Meanings
 
AI7 W11    Terminology Data Categories Survey
 
AI7 M12    Minutes of Oak Ridge Meeting
 
AI7 W13    Preliminary TC 37 SC3 report
 
AI7 W14    Myrdal working paper (based on Hull paper, AI7W14 Ver01)
 
AI7 A15    Agenda for the Vienna Meeting
 
AI7 W16    Nested and Flat DTDs, 1991-11-03
 
AI7 M17    Minutes of the Vienna Meeting
 
AI7 D18    Draft for the AI7 contribution to P2
 
AI7 D19    AI7/P2 DTDs
 
AI7 W20    Aslib paper
 

*************************************************************************
Attachment #4: ISO/TC 37/SC 3 N 89 en Rev 2
                                                                                       1991-11-19
Re: Item for future work
Proposer: Secretariat of ISO/TC 37/SC 3
Result of discussion at the 8th meeting of SC 3, Vienna, 1991-11-18
 
Draft Title
 
Computational aids for terminology ~ Formats for data interchange ~ SGML
application for terminological entries
 
Scope
 
The standard provides a universally applicable format for terminological data
based on SGML designed primarily for interchange purposes.
 
Purpose and justification
 
Specific aims and reason for standardization activity
 
The project will utilize the SGML-reference concrete syntax provided in ISO
8879 and will standardize one or more document type definitions (DTDs)
applicable to terminological entries.
 
The project will determine whether there is a need to standardize more than
one DTD in order to accommodate all possible types of term entries.
 
This SGML support facility will be used for terminology management and
interchange.  It will be designed to facilitate interaction with:
*   terminological databases
*   lexicographical databases
*   text bases
*   thesaurus and documentation databases
*   bibliographical databases
*   other SGML applications.
 
Main interests
 
Such an activity will ensure coordination with ISO efforts regarding
information technology (IT) applications in standards development and
publishing in the terminology area.  This activity conforms to the
Recommendations of the ISO/STACO seminar "Faster and Better: IT applications
in standards development" of 1991-06-3/4, which were subsequently endorsed by
the ISO Council at their Madrid meeting of October, 1991:
    1.3 Terminology data, being today derived from various sources and updated
    in various organizations, should be standardized world-wide in order to
    facilitate access, readability and mergeability with other dictionary-like
    data.
Potential users of this standard include all individual and institutional
producers and users of terminological information.
 
Feasibility, urgency, and benefits for international harmonization
 
Support software for SGML processing has only recently become available.
National and international projects in the field of information and
documentation have been initiated with the objective of standardizing SGML
formats for special document types, e.g. standards, journal articles, etc.
 
Early participation on the part of the international standards community will
contribute to the development of a harmonized format for the international
interchange of scientific and technical documents.
 
Project stages (not necessarily in sequential order)
 
0   Preliminary studies
    *   relationship to ISO 6156 MATER and other terminological interchange
        formats
    *   relationship to other SGML applications
    (Proposal attached, see Annex 1)
 
1   Terminological questions
    *   related terms not included in ISO 8879
    (Proposal attached, see Annex 2)
 
    Need for coordination with
    *   ISO/IEC JTC 1/SC 1 "Information technology; Vocabulary"
    *   ISO/TC 46/SC 3 "Terminology of information and documentation"
        [.......]
 
2   Identification of mandatory and optional data elements that should be
    incorporated in the standardized SGML terminology format
    (Proposal attached, see Annex 5)
 
3   Development of the standardized SGML format for terminological entries
    according to ISO 8879
    (Proposal attached, see Annexes 3, 4, 6)
 
    Need for coordination with
    *   ISO/IEC JTC 1/SC 18 "Text and office systems"
 
4   Support of character sets
 
    Need for coordination with
    *   ISO/IEC JTC 1/SC 2 "Character sets and information coding"
    *   ISO/TC 46/SC 2 "Transliteration" and SC 4/WG 1 "Character sets"
    (Proposal attached, see Annex 7)
    The SGML format shall support the bi-directional transliteration of
character sets.
Project Allocation
 
The project will be allocated to ISO TC 37/SC 3/WG 2, "SGML applications"
(Convener: Dr. Budin, N).
 
Time Schedule
 
NP      92-02      (Ballot to close 92-05)
WD      92-04      (Circulation for discussion at 9th meeting)
WD      93-04      (Circulation for decision at the 10th meeting of SC3 whether
                   to circulate as a CD)
CD      93-11      (Ballot to close by 94-02)
                   Forward text to SCC for French translation
DIS     94-11      (Ballot closes: 95-05)
                   Resolution of comments at the 11th meeting of SC 3
IS      95-11
 
 
 
 
 
 
 
 
************************************************************************               Draft
Attachment # 5, TEI AI7 M17
 
Demonstration files: MMUTS, NORM utilities                         1991-11-16
 
The TEI-TERM documents A, B, C were generated from a short sample MicroMATER
file. A is the flat document generated directly from the MM file.  B is the
normalized (nested) document generated from the flat document.  C is the
normalized document generated directly from the MM file.
 
It should be noted that the MMUTS and NORM utilities used to generate these
files (31.10 and 01.11.91, resp.) represented an older version of the TEI-TERM
DTD that did not parse properly.  This version of the DTD placed the </term>
tag at the end of the <tig> instead of directly following the term.  This error
has been corrected in the output files.
 
MicroMATER file
 
 
{mm} 2
{ln} en-pt
-----
*day
{1def} nao ^B noite
{defd} 9 Oct 91
{1ltu} dia
{LTS} pre
{0def} not night
*night
{1} noite
{LTY} syn
{def} nao ^B dia
{1defs} 123ABC
{DE:2LTU} Nacht
{def} nicht Tag
{rl:src} 789XYZ
{rl:srcd} 15 Oct 91
{rl:rst} wor
@!
 

A.  Flat TEI-TERM document
 
<!DOCTYPE TEI-TERM>
<TEI-TERM>
<header>
</header>
<text>
<front>
</front>
<body>
<termEntry>
    <term lang=en group=0> day </term>
    <descrip type=definition group=1 n=1> nao ^B noite </descrip>
    <admin type=date group=1 depend=1> 9 Oct 91 </admin>
    <term lang=pt group=1> dia </term>
    <admin type=preferredTerm group=1> </admin>
    <descrip type=definition group=0> not night </descrip>
</termEntry>
<termEntry>
    <term lang=en group=0> night </term>
    <term type=synonym lang=pt group=1> noite </term>
    <descrip type=definition group=1 n=1> nao ^B dia </descrip>
    <citnRef target=123ABC group=1 depend=1>
    <term lang=DE group=2> Nacht </term>
    <descrip type=definition group=2> nicht Tag </descrip>
    <termEntryLevel type=citnRef target=789XYZ n=2>
    <termEntryLevel type=admin type=date depend=2> 15 Oct 91 </termEntryLevel>
    <termEntryLevel type=admin type=termEntryStatus> workingEntry
</termEntryLevel>
</termEntry>
</body>
</text>
</TEI-TERM>
B.  TEI-TERM nested document generated from flat document A using the NORM
utility
 
<!DOCTYPE TEI-TERM>
<TEI-TERM>
<header>
</header>
<text>
<front>
</front>
<body>
<termEntry>
    <tig>
        <term lang=en> day </term>
        <descrip type=definition> not night </descrip>
    </tig>
    <tig>
        <term lang=pt> dia </term>
        <descrip type=definition> nao ^B noite
            <admin type=date> 9 Oct 91 </admin>
        </descrip>
        <admin type=preferredTerm> </admin>
    </tig>
</termEntry>
<termEntry>
    <termEntryLevel>
        <citnRef target=789XYZ>
            <admin type=date> 15 Oct 91 </admin>
        <admin type=termEntryStatus> workingEntry </admin>
    </termEntryLevel>
    <tig>
        <term lang=en> night </term>
    </tig>
    <tig>
        <term type=synonym lang=pt> noite </term>
        <descrip type=definition> nao ^B dia
            <citnRef target=123ABC>
        </descrip>
    </tig>
    <tig>
        <term lang=DE> Nacht </term>
        <descrip type=definition> nicht Tag </descrip>
    </tig>
</termEntry>
</body>
</text>
</TEI-TERM>
C.  TEI-TERM nested document generated directly from the MM file using the MMUTS
    utility
 
<!DOCTYPE TEI-TERM>
<TEI-TERM>
<header>
</header>
<text>
<front>
</front>
<body>
<termEntry>
    <tig>
        <term lang=en> day </term>
        <descrip type=definition> not night </descrip>
    </tig>
    <tig>
        <term lang=pt> dia </term>
        <descrip type=definition> nao ^B noite
            <admin type=date> 9 Oct 91 </admin>
        </descrip>
        <admin type=preferredTerm> </admin>
    </tig>
</termEntry>
<termEntry>
    <termEntryLevel>
        <citnRef target=789XYZ>
            <admin type=date> 15 Oct 91 </admin>
        <admin type=termEntryStatus> workingEntry </admin>
    </termEntryLevel>
    <tig>
        <term lang=en> night </term>
    </tig>
    <tig>
        <term type=synonym lang=pt> noite </term>
        <descrip type=definition> nao ^B dia
            <citnRef target=123ABC>
        </descrip>
    </tig>
    <tig>
        <term lang=DE> Nacht </term>
        <descrip type=definition> nicht Tag </descrip>
    </tig>
</termEntry>
</body>
</text>
</TEI-TERM>