Article:		To Tag or Not to Tag
Subject:		CyberHamlet
Date:		2004-06-23 01:31:19
From:		ben_
For my PhD on several english an german Hamlet Editions, I work with an xml appraoch, to use mutliple annotations of one and the same primary text. We developed this technique for a different project [hytex] and a different corpus, but I worked pretty well: To get use of different annotations, we unify the XML-Data and write modularised XSLT-Stylesheets wo work on this tagsoup. This was s hard and frustrating in the beginning, but generally we can devide the diffrent Layers via the mode-Attribute quite well. I think that using, multiple annotation for variorum edition or other highly-metadata annotations is a great way to keep HIGHLY complex XML-Documents editable in sourcecode. The only existing problem is, the the software for unifing different Annotations is [a later] development stage, not not working perfect yet. Another interesting point, we used in the hytex-project and that will be added to the Hamlet-corpus is a crossreference-mechanism via a semantic network[XML-Topic Maps], which is build up besides the textual corpus. This approach give you the chance to model central-Informations on the domain data-oriented and not document-oriented. And by linking the sematic network to all variations of a play, you get rid of the problem of redundant meta data in different editions of on play. And a third approach we follow is not only to markup struktural and presentational data, but to create another layer with a narratological informations. This is especially usefull in our case, as we try to setup a synopse of 18 german Translation, which differ in some case enourmously from the original play. So you additionally have the chance not only to compare similar scenes and speeches, but to compare similar events, and different "event-architectures" of different editions.

Previous Message

Next Message