Closed behas closed 12 years ago
Implemented via http://wikipedia-miner.cms.waikato.ac.nz/
Abstracts/descriptions still need to be fetched via DBPedia's JSON(?) API:
e.g., http://dbpedia.org/data/Chile.json
"http://dbpedia.org/ontology/abstract" is the key; we only need the english abstract.
A standalone installation of WikipediaMiner is now available at: http://samos.mminf.univie.ac.at:8080/wikipediaminer/
Please adapt the request URI accordingly.
Fixed the issue with 2 different implementations:
On my machine, the SPARQL method worked significantly faster. I commented out the other method just in case.
Cut off text after 295 characters and added "(...)" to the end to control text length.
Implement a lookup service adapter that finds DBPedia URIs for some given annotation text.
Further details are here: http://wiki.dbpedia.org/spotlight/usersmanual?v=i0m