SIG:Ontologies

Introduction
In May and June 2004, there was a discussion on the TEI mailing list about prosopographical tags. This lead to a suggestion that detailed information about persons (physical and legal), dates, events, places, objects etc. and their interpretation could be marked up outside the text, and that this could be connected in on-going ontology work being done e.g. in the Museum community, such as the Conceptual Reference Model (CIDOC-CRM). The result of this was the establishment of a Ontologies SIG at the 4th annual members meeting of TEI in October 2004.

The SIG runs a mailing list on this topic. To join, send a message to tei-ontology-sig-request@lists.sourceforge.net with the word SUBSCRIBE in the header or the body of the email.

The SIG is convened by Øyvind Eide, of the Unit for Digital Documentation at the University of Oslo, and Christian-Emil Ore, of the Unit for Digital Documentation at the University of Oslo, who is also the chair of The International Committee for Documentation of the International Council of Museums (ICOM-CIDOC).

Meeting on October 23 2004
The meeting was held during the 4th annual members meeting of TEI in Baltimore in October 2004.

As a motivation for the conveners' interest in the topic, a short introduction to the CIDOC-CRM was given, together with a description of how the Unit for Digital Documentation use the standard. This was combined with a lively discussion about TEI and ontologies in general, and on the CIDOC-CRM standard in specific.

In 2004 and 2005, the conveners have promised to do the following work:


 * Set up a mailing list (2004)
 * Create a simple web page (2004)
 * Locate people interested in participation
 * Communicate with ICOM/CIDOC
 * Do TEI-CRM work on some of our data
 * Report on our work to the SIG list
 * Report progress at the next TEI meeting

In addition, we hope others will report on their work to the mailing list.

It was also agreed that the following tasks should be done by someone, but the responsibility was not appointed to anyone in specific:


 * Identify all elements in TEI with ontological relevance
 * Define mappings from these elements to CRM and to other ontological systems
 * Identify missing elements in both standards
 * Find the borders between TEI and ontologies in order to avoid duplication
 * Investigate the use of METS as a way to connect TEI to CRM
 * Feedback into TEI P5 based on the aforementioned tasks
 * Suggest more complete solutions on the integration/connection issue

Open meeting on June 15 2005
The TEI Ontologies SIG held an open meeting during the ACH/ALLC Conference in Victoria, Canada, on June 15, 2005. Six persons were participating.

Introduction
In the minutes from the lest meeting, two groups of tasks were set up:

1. Tasks the conveners promised to do in 2004 and 2005

2. Tasks we agreed should be done, but with no appointed person responsible

The tasks in group 1 are all on-going or finished, while of the tasks in group 2, only the identification of elements in TEI with "ontological relevance" (more on this expression below) is started.

Motivation
As a motivation for his interest in this work, Øyvind Eide described a project in which an analysis of relationships between person name elements in a TEI document was modeled in CIDOC-CRM. This work was discussed by the group, together with some aspects of the CIDOC-CRM model.

The group agreed on the general idea that there are things, such as relationships between persons, that should be modeled outside TEI, but with a possibility for links to TEI documents.

Identification of TEI element
This item was presented to the group as "Identifying TEI elements of special ontological interest". This initiated a discussion, because the group rejected the wording. Several other phrases were suggested to replace "special ontological interest", among them "extra-textual (ontological) interest" and "references to the physical world". None of these truly covers what we want to express, though.

Nevertheless, there were an common general understanding about what kind of elements we were talking about: Elements such as names, date and performances of plays, while elements such as italics, stanza and paragraphs are outside the scope of this SIG.

Time did not permit any actual investigation into which elements in TEI are of interest.

The on-going work in the FRBR/CIDOC-CRM Harmonization Group was discussed. It was argued that the TEI community should be involved in this work.

2005-08-01 Øyvind Eide

Meeting on October 28 2005
The meeting was held in Sofia on the morning of October 28, 2005. Six persons were participating.

With reference to the last meeting in Victoria, there was a discussion about the scope of the SIG. We have not been able to make any precise definition, and will have to do with the more practical description of our scope as "elements such as persons, places, dates and events".

The SIG agreed that the example presented at the first day of the TEI meeting by Conal Tuohy at the New Zealand Electronic Text Centre is a good example of the type of work interesting for this SIG.

The group discussed whether we should propose a generalized event element in TEI, similar to the name element. We did not conclude on this, but it should be further discussed.

We also discussed different ways to connect a TEI document to ontological information. A TEI document can contain the ontology, either as reference or included. We will work further on examining practical implementations of such connections.

Plans for 2006
We feel that we should now be able to do practical work to see how the ideas discussed in the SIG will work in practical implementation. The conveners will work on this, and will publish results, both on the SIG list an WIKI, and through other channels.

It was also interest in exploring ways to include Topic Maps in the TEI header.

The conveners will publish examples of our work on the WIKI, as will the New Zealand Electronic Text Centre. We hope this can lead on to a best practice document.

The SIG will probably have a meeting at the Digital Humanities conference in Paris in 2006.

2005-12-09 Øyvind Eide

Meeting on October 28 2006
11 persons were present at (parts of) the meeting.

The last year
The activity on the list and the WIKI has been quite limited the last year, and only the convener have posted anything either place. There is a visible interest in the topic of the SIG in many areas, though.

The annual conference of The International Committee for Documentation of the International Council of Museums (ICOM-CIDOC) was held in Gothenburg September 10-14. During the meetings of the Documentation Standards Group, liason with other parties was discussed and the work in this SIG came up. The group expressed support for the work done by the Unit for Digital Documentation at the University of Oslo on the relationship between CIDOC-CRM and TEI, and hoped the work would continue. This support includes, of course, to the work being done by other members of this list and participants at the SIG meetings as well. The CIDOC do not have an economy which makes it possible with financial support, but moral support is also good!

General discussion
The background of the SIG was presented together with a discussion of the difference between TEI and CIDOC-CRM.

The problem with authority control came up, how the problem relates to the possible use of FRBM and CIDOC-CRM in connection with TEI was discussed. There is a need for authority services, and the development and use of such services is related to the ontology work.

The cut-off point between TEI and conceptual systems/ontologies is an interesting problem that has to be continously worked on. How much information should be stored in TEI, how much in other formats with pointers to the TEI document?

Ontological information systems will often want to point to well defined access points in TEI.

CIDOC-CRM
The draft for a CIDOC-CRM mapping of some TEI P5 elements being done by EDD will be publised on the mailing list and the WIKI this year. The possibility for use of the equiv element to express such a mapping will be examined. Ãyvind Eide and Christian-Emil Ore is repsonsible for this this.

We will also work on the construction of parts of a system for automacially building CIDOC-CRM comaptible models based on TEI documents and the TEI-->CIDOC-CRM mapping in 2007.

FRBR
Gregory Crane at the Perseus Project will do some work on the relationship between FRBR and TEI in 2007 and report to the list and the WIKI.

TEI
Instances of persons are just added to TEI at element level, and place may be on the way (as opposed to names of persons and plases). There are event elements, but only specific (persEvent, event in transcription of speach). Should a general event element be added as well?

And maybe a general object element?

There have been some work on different practical methods to relate ontological information to TEI documents. More experiments are welcome.

Meetings in 2007
The next meeting at the TEI meeting in October 2007. There may also an interim meeting at the Digital Humanities conference in June 2007 if potential participants are interested.

2006-11-21 Øyvind Eide

Papers
Some paper with relevance to the work of this SIG have been presented lately. The following is a short list, please add more!

[http://www.edd.uio.no/artiklar/teknikk_informatikk/CIDOC2006/EIDE_HOLMEN_Reading_Gray_Literature.pdf Eide, Ø and Holmen, J: Reading Gray Literature as Texts. Semantic Mark-up of Museum Acquisition Catalogues]

[http://www.allc-ach2006.colloques.paris-sorbonne.fr/DHs.pdf Eide, Ø and Ore, CE: TEI, CIDOC - CRM and a Possible Interface between the Two. Page 62-65]