Hello, mo wrote: > Hi, > > thanks for the reply. > >>> * We can easily get >15000 projects from freebase [1], even along >>> with >>> Wikipedia keys, but which of the multiple keys are the ones used in >>> dbpedia? Is there any fixed rule/heuristics which to use? >> What exactly do you mean by "key"? DBpedia URIs correspond to >> Wikipedia >> articles, i.e. http://dbpedia.org/resource/$something contains >> information about the object described at >> http://en.wikipedia.org/wiki/$something. We took care not to change >> the >> Wikpedia article URIs. >> >> I don't know how Freebase handles it. There was some discussion that >> the >> Freebase guys want to publish their data as Linked Data and >> interlink it >> with DBpedia, but as far as I know that has not happened yet. Freebase >> does have dumps, which (they claim) can be easily converted to RDF. >> That >> might be a starting point. > > OK, so I will ask the freebase mailing list how they handle this.
Please ask them to go for Linked Data and DBpedia interlinkage. :-) >>> * The dbpedia data on buildings seems richer - but I can't find out >>> how to retrieve ALL architectural projects in dbpedia. I tried [2] >>> athttp://dbpedia.org/sparql >>> , but it stops athttp://dbpedia.org/resource/Beijing_Zoo >> There is a limit of 1000 results on the SPARQL endpoint. So either you >> have to use limit and offset or you just download the complete YAGO >> file >> and grep for http://dbpedia.org/class/yago/Building102913152. The data >> set is called "YAGO classes" and can be downloaded from the DBpedia >> download page [1]. > > OK, thanks. I found 12.627 buildings in the Yago classes file, quite > nice. I just noticed that you don't get all buildings by just grepping for http://dbpedia.org/class/yago/Building102913152. There are also various subclasses of building, which have instances assigned to them. To get all those you can add "define input:inference 'http://dbpedia.org/resource/inference/rules/yago#'" (without ") at the beginning of your SPARQL query in your previous mail (this enables Virtuoso subclass inferencing) and then use LIMIT and OFFSET to find all solutions in several queries. You'll get 28582 buildings this way. >>> * Also, I find it hard to find out which properties buildings >>> typically have in dbpedia. How do you do that? Just inspect lots of >>> instances? Any tricks there? >> Basically yes. They are available as Linked Data, e.g. you can just >> call >> http://dbpedia.org/resource/Beijing_Zoo in your browser. The available >> topics are determined by Wikipedia and in particular the infoboxes. >> The >> Beijing_Zoo page [2] does not have an infobox, so you won't get much >> building specific information there. You could watch out for common >> Wikipedia infoboxes related to buildings (if those exist) to find >> typical properties. > > I feared so :) Thanks! The YAGO class hierarchy itself is also an interesting source of information (in particular in the upcoming DBpedia release), i.e. the class name alone may tell you something about a building. We will also interlink the classes with YAGO using owl:equivalentClass, so you may get additional information from there (soon). Kind regards, Jens -- Dipl. Inf. Jens Lehmann Department of Computer Science, University of Leipzig Homepage: http://www.jens-lehmann.org GPG Key: http://jens-lehmann.org/jens_lehmann.asc ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Dbpedia-discussion mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
