| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

Travel narrative group

Page history last edited by Øyvind Eide 12 years, 3 months ago

Problem identification

 

Place names in travel narrative, marked up, text encoding.

 

Tagging of place names - can be identified, hard to locate them, cannot put coordinates directly in.

 

Two types of connections: 1) method in different projects, 2) Linking place names.

 

There are links between place names if the travel narrative describes travel in the same areas, at least from roughly the same time perios

 

2) Using coordinates is a hypothesis, is problematic to use for interconnection. But useful for search.

 

Method: i) mark up place names, ii) connect to gazetteer and/or other sources. The name authority should provide coordinates.

 

Polygons are dangerous, not meaning then.

 

Use administrative areas and places, linked, but still separate.

 

Automatic place name tagging difficult for early modern text, cleaning up becomes too expensive. Manual tagging works. Crowdsourcing?

 

Use Word styles to encode place names, convert to TEI; not crowdsourcing but community sourcing.

 

Detect place names is one thing, actually identifying them is another thing.

 

10-20 travel books marked up with place names, what next?

 

Detailed markup at pargraph level, or lower? Seeing travel narratives a linear, or not? Intertext, linking references with places. Want more semantic tagging, but expensive.

 

CS people try to mine slightly later (19th centory) resources now, will have to see the results; agitating travelling politicians.

 

Automating sematnic markup is very different, but identifying keywords?

 

Context is necessary to identify place names.

 

Patterns of travel in travel narrative?

 

Scenario:

 

1) 20 travel books with place names manually marked up.

 

2) Coordinates for many of the places are found by external sources.

 

3) Give candidate coordinates for the ones not identified. Look at how they occur in different books, look at context weighted by text distance in several books. Create areas of possible locations, as an input for identification.

 

The problem is to identify that names refer to the time place.

 

Another problem is places mentioned and not visited - the problem of functional encoding of place names.  RDF - interconnecting places. Event based structure on top of it.

 

This is meant for qualitative studies, not quantiative, so statistical error as the social historians have is problematic.

 

Itineraries vs. travel descriptions, not a clear distinction.They could be quite usefuld

 

Travel guides are also useful.

 

Hard to draw any conclusions. Need more work to clarify the methods to be used.

 

Need to dig deep into the material in order to try things in practice, both autmated markup, crowd/community-sourcing, and identifcation of place names. Need to build stuff to see how they work, to get any further.

 

Implementation

 

How do we link to the resources?

 

Walled garden, easy to link inside.

 

Trying manually to link some Norwegian place names: http://www.oeide.no/Pelagios/linkTest.html

 

We can only ask questions we already know the answer to. Impossible to link these resources based on searching.

 

Gjesåsen kirke is a small, insignificant church, but many sources mention it; problem is, even manual linking of them is pretty difficult. Quite helpful to know things like the changes to the municipality of which this church is a part.

 

Deep linking, in any sources found: only in Rygh and Wikipedia.

 

Google is not very useful either. Mostly irrelevnt stuff.

 

Nothing on Egge's long history in Wikipedia (as of today; could be different tomorrow, of course).

Comments (0)

You don't have permission to comment on this page.