Difference between revisions of "Annotation and indexing"

Latest revision as of 15:01, 31 October 2018

In a digital edition, annotation can be provided in several ways and at several levels. The most simple form of annotation are notes anchored to a location in the text. It is also possible to hyperlink certain stretches of text (e.g. those that refer to persons or other named entitites to external information, created by the project or elsewhere. Annotation can also be provided at a higher level, for instance about a chapter or a letter in a collection. Most projects will also have some level of annotation at the level of the entire work or collection being edited, for instance about its publication history or its literary context.

Often, it will be desirable to create annotations of multiple types, e.g. in order to filter the display of annotations by type. This can occur at the levels of annotation.

Indexing is related to annotation in that it can be used to link stretches of text to an index term, often provided with explanations.

Local annotations

Local annotations are annotations anchored to a location in the test. We assume here that a point in the text is sufficient and that there is no need to mark a stretch of text as the target of annotation (but see Indexing, below).

In the TEI it is possible to create 'embedded' notes, where the note element is contained within the transcription at the place where it applies. The disadvantage of this approach is that is diminishes the legibility of the text for the encoder. Therefore, at Huygens ING the preffered method for local annotation is to create an anchor in the text that identifies the place where a note should be inserted. The actual notes are stored in a separate <div> in the same file.

In order to create an annotation, we place an empty <anchor/> element within the transcription. The anchor should contain a @n attribute with a short keyword and a @type="note". A corresponding <note> element should be inserted in the <div type="notes"> element elsewhere in the document. The @n must have the same value in both elements. The notes div can contain subdivisions by language and further subdivisions based on the type attribute of the note. Within each subdivision, the notes should be arranged based on their location in the text. For example:

<p>This is an example<anchor n="mynote" type= "note"/> of how to create a note</p>

<div type="notes">
 <div xml:lang="en">
  <note n="mynote">This is the note</note>
  … other notes in English…
 </div>

 <div xml:lang="nl">
  <note n="mynote">Dit is de noot</note>
  … other notes in Dutch …
 </div>
</div>

Ongoing topic-annotations

In a correspondence edition, ongoing topic-annotations refer a reader back to an earlier letter where a recurring topic in the correspondence was discussed and explained. Ongoing topic-notes are stored in a division of type 'ogtnotes'.

For example:

<div type="ogtnotes">
   <note type="ogt" n="1">Mondriaan's werkzaamheden in Rotterdam <ref target="1916XXXX_SLIJ_1364.xml">...</ref>.</note>
</div>

The example assumes that the ongoing topics are numbered (n=1). The reference is to the first letter where the topic is discussed.

See also

• Transcription: annotations (MVN)
• Transcription: annotations (Mondrian)
• XML Huygens