search

linkedtv / editortool

all issues concerning the LinkedTV EditorTool
2 stars 1 forks source link

Extend DBPedia lookup to other languages #1

Open lyndonnixon opened 10 years ago

lyndonnixon commented 10 years ago

DBPedia lookup is only for the English DBPedia so I couldn't add Dutch or German concepts unless I copy and paste the already known URI

jblom commented 10 years ago

It seems the DBpedia lookup API does not support searching in other languages. In these cases the SPARQL end-point must be used, which would probably be too slow for autocompletion...

http://stackoverflow.com/questions/6747019/how-get-dbpedia-data-in-a-specific-language

Currently not sure how to solve this.

rtroncy commented 10 years ago

Use another lookup service. I think @RubenVerborgh developed one as well. Another alternative: http://search.datao.net/ which has indexed the 122 endpoints from SPARKLES. There are many other services. I think you should not close issue that you don't fix now.

lyndonnixon commented 10 years ago

I havent used this but http://wiki.dbpedia.org/Downloads2014 offers datadumps that I understood to be just the titles of articles in the German and Dutch DBPedias

rtroncy commented 10 years ago

This is another option of course, although this requires some serious engineering effort (the dumps are very large!) ... and my point is that many people have the same need and went to the same process.

jblom commented 10 years ago

Thank you both. I'll first check out if there is a suitable online service (I'll ask Ruben if he knows something as well), indexing a dump is something I'll definitely try to avoid.

RubenVerborgh commented 10 years ago

Hi all, I didn't develop a DBpedia lookup service; just made a client of the existing service.

rtroncy commented 10 years ago

@RubenVerborgh I do remember to have seen a demo on your site where I could easily look for a dbpedia URI using a query string. Did I dream this? I don't think it was connected to Spotlight

RubenVerborgh commented 10 years ago

@rtroncy I probably wasn't my site… I cannot recall this. What we have done, is document reconciliation for OpenRefine, but that one relies on public SPARQL endpoints, which might not be the most stable option.