Difference between revisions of "Darmstadt Bootcamp/AXLD"

From IntereditionWiki

(Tasks: update)
(updated and added link to GitHub)
 
Line 10: Line 10:
  
 
===Technology===
 
===Technology===
* Linked Data
+
* Linked Data: SPARQL, DBPedia
** SPARQL
+
* XML family: XPath, XSLT
** DBPedia
+
* Open Annotation Collaboration: my own naive implementation
** ...
+
* Ruby: Rails, RDF.rb, Spira, ruby-xslt, sparql-client
* XML family
+
** XPath   (XPointer??)
+
** XSLT
+
* Open Annotation Collaboration  
+
** implement our own?
+
** ...?
+
* Ruby  
+
** Rails
+
** rdf.rb
+
** Spira
+
** ...
+
  
 
===Tasks===
 
===Tasks===
Line 50: Line 39:
 
# Given a URI, harvest open annotation links from a predefined or given list of sparql endpoints, and return a list of annotation URIs
 
# Given a URI, harvest open annotation links from a predefined or given list of sparql endpoints, and return a list of annotation URIs
 
# Given a URL to an XML (e.g. TEI) document and a list of annotation URIs, incorporate the annotations and return the integrated XML.
 
# Given a URL to an XML (e.g. TEI) document and a list of annotation URIs, incorporate the annotations and return the integrated XML.
 
+
===Demo Code===
 +
[https://github.com/abartov/raxld RAXLD is on Github]
 
===Contact===
 
===Contact===
 
* [[User:Asaf|Asaf Bartov]]
 
* [[User:Asaf|Asaf Bartov]]

Latest revision as of 12:11, 4 March 2011

(up to Darmstadt Bootcamp)

This is the page for the Annotate XML with Linked Data (AXLD) tool.

Goal

  1. Parse XML for annotation-worthy strings (Proper-case words/phrases at first?)
  2. Collect Relevant Links and Multimedia from the Semantic Web
  3. Present suggested links and allow selection (one by one or by include/exclude rules)
  4. Incorporate selected links/media as annotations to source XML

Technology

  • Linked Data: SPARQL, DBPedia
  • XML family: XPath, XSLT
  • Open Annotation Collaboration: my own naive implementation
  • Ruby: Rails, RDF.rb, Spira, ruby-xslt, sparql-client

Tasks

  • Pick sample TEIs -- DONE
  • Set up a "texts server" that renders TEIs into HTML -- DONE
  • Set up a Linked Data source publishing basic OAC annotations -- DONE
  • Set up a harvest process -- DONE
  • Prepare annotations that target specific locations in the sample TEIs -- DONE
  • Make the text server pre-process the TEIs before rendering them, interpolating the annotations into the XML before sending it to the XSLT processor. -- DONE
  • Add to the TEI->HTML XSL some handling for "annotations" -- DONE
  • Prepare a nice CSS for annotations -- DONE
  • add support for separate layers of annotations
  • present more data per annotation (annotator FOAF, etc.)

Auto-suggester

  • Identify "interesting" terms
    • Just any word with a capital letter -- DONE
    • Something more clever
  • Query DBPedia
    • Retrieve abstracts for interesting terms
    • Present suggestions
    • Allow select/discard en masse
  • Add annotations to the annotation repository
    • Append to the triple store

Exposure as a service?

  1. Given a URI, harvest open annotation links from a predefined or given list of sparql endpoints, and return a list of annotation URIs
  2. Given a URL to an XML (e.g. TEI) document and a list of annotation URIs, incorporate the annotations and return the integrated XML.

Demo Code

RAXLD is on Github

Contact