From: Conal T. <Con...@vu...> - 2004-12-06 22:54:08
|
=D8yvind Eide: > This will not be a very high traffic list, I would suppose. From the > conveners' side, we will post when we have questions we would like to > discuss, or to inform the members of interesting articles or work > reports. If anyone else wants to start a discussion, please do! OK, I'll kick off :-) Who is using TEI in conjunction with ontologies, and how are you doing = it? Briefly, at the NZETC we are starting to use TEI with the CIDOC CRM (a = museum ontology), using Topic Maps as our implementation technology. We = are also using MADS for authority metadata (canonical names etc. of = people, places...) The linkages from the TEI are primarily encoded using the TEI <name> = element, with key attribute pointing at external topics. The nature of = each such linkage is normally defined by the parent element of the = <name> element. e.g. <author><name key=3D"blah">Mr B. = Blah</name></author> would produce an "authorship" association between = the person "Mr B Blah" and the TEI document. We are not presently using any of the more directly ontological features = of TEI like feature structures and taxonomies. Cheers Con |
From: Christian W. <wi...@ka...> - 2004-12-07 03:06:18
|
"Conal Tuohy" <Con...@vu...> writes: > =D8yvind Eide: > >> This will not be a very high traffic list, I would suppose. From the >> conveners' side, we will post when we have questions we would like to >> discuss, or to inform the members of interesting articles or work >> reports. If anyone else wants to start a discussion, please do! > > OK, I'll kick off :-) > > Who is using TEI in conjunction with ontologies, and how are you doing it? We are using TEI with a homegrown ontology for the Tang Knowledgebase Project here in Kyoto. We do quite similar things (e.g. linking from <name> and other elements) and are also using TopicMap technology. We are working with texts, rather than bibliographies, so the association is in most cases not immediately derived from the context. At the moment, one of the headaches is a software environment for maintaining this stuff and so we are at the beginning of developing our own stuff with eXist and some Java libraries. I would be very interested to hear what kind of s/w environment other projects are using. We did try the OKS Suite from Ontopia, but it choked on some of the larger topicmaps. All the best, Christian --=20 Christian Wittern=20 Institute for Research in Humanities, Kyoto University 47 Higashiogura-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8265, JAPAN |
From: Peter B. <pb...@xs...> - 2004-12-07 10:21:48
|
> Who is using TEI in conjunction with ontologies, and how are you doing it? I'm trying to encode aspects of interpretation of literary works. There are ptr and xptr elements in a TEI-encoded XML file, to which I link from an RDF file. The ontology is developed ad hoc, based on the kind of things I want to describe in the work at stake. The ontology at present is created as an RDF Schema. Both the ontology and the RDF instance file are created more or less manually in an XML editor. Querying is done through a php application that uses the RDF API for PHP; for performance reasons the RDF(S) data are being stored in a MySQL database. Especially in creating the instance file I will need software support. The software will have to be aware of the ontology, but it should also somehow have to know what elements in a TEI file are the elements that I want to annotate. And it should automatically add derived triples (my annotations are annotations of fragments of text; I want to generate corresponding annotations on the higher level textual units that contain these fragments). Best regards, Peter |
From: E. <oyv...@mu...> - 2004-12-08 15:02:11
|
On Tue, Dec 07, 2004 at 11:53:59AM +1300, Conal Tuohy wrote: > Who is using TEI in conjunction with ontologies, and how are you doing it? Our main work on encoded text and ontologies is based on printed museum catalogues. These catalogues, from archaeological and ethnographic museums, were printed as annual "aquisition catalogues" from the 19th century to late 20th century. When we digitized them in the 1990s, we did the digitizing by scanning/OCR reading and SGML tagging of the texts in the printed catalogues. The tagging was based on a local DTD developed in the early 1990s. We are currently importing the data into an event oriented, CIDOC-CRM based database. This means that we are doing a mapping from the SGML tagged text to CIDOC-CRM. While this is not directly TEI work, the process would have been quite similar if the SGML data had been encoded in TEI. There is a short description of the work in this abstract: Holmen, J: "From an XML tagged acquisition catalogue to an event based relation database" at the DRH 2004 conference, URL: http://drh2004.ncl.ac.uk/abstract.php?abstract=168 There is also an article forthcoming in the proceedings of the CAA Conference due to be published next spring. I will notify the list when it is out. Apart from that, we are also working along similar lines as Conal Tuohy, Christian Wittern and Peter Boot, but it is in quite an early stage. I hope to be able to report more on that in a few months! Kind regards, Øyvind Eide |