Some bugs on dbpedia...

17 views
Skip to first unread message

Axel Polleres

unread,
Jan 4, 2012, 2:19:07 PM1/4/12
to Sören Auer, Kingsley Idehen, Chris Bizer, pedant...@googlegroups.com
Dear Chris, Soeren, Kingsley,

FWIW, and FYI, I posted some XML and literal-typing bugs worthwhile being mentioned to the pedanticweb group at
https://sourceforge.net/tracker/?func=detail&aid=3469498&group_id=190976&atid=935520

Details there, just for explanation: I totally do have understanding for strange/bogus triples resulting from automatic extraction from wikipedia's infoboxes
and I certainly appreciate the tremendous efforts you and your groups have made to publish that data over the past years, great job really!

However, I kindly ask you to follow two small things in your published data on dbpedia:
a) please check your exporters to produce/publish valid RDF/XML only
b) please do not publish ill-typed literals, i.e. literals with a lexical value out of the lexical space for XSD datatypes (e.g. xsd:date values in xsd:gYear typed literals)
since both these issues let common RDF tools choke.

Thanks a lot & cheers,
Axel


--
Dr. Axel Polleres
url: http://www.polleres.net/ twitter: AxelPolleres

Kingsley Idehen

unread,
Jan 4, 2012, 2:44:48 PM1/4/12
to Axel Polleres, Sören Auer, Chris Bizer, pedant...@googlegroups.com
Happy New Year!

Nice time to ping we'll collectively see what we can do.

--

Regards,

Kingsley Idehen
Founder& CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Peter DeVries

unread,
Jan 12, 2012, 8:19:03 PM1/12/12
to pedant...@googlegroups.com
I have been thinking about trying to help with these efforts.

I have noticed some things with the data that does not really have to do with the DBpedia team but about people putting in the wrong "kinds" of things into the different properties. For instance, a binomial name should just have two words, but some people put in the trinomial name (for a subspecies) in that taxobox field.

I am wondering if people would like stable URI's for the various clades (taxon levels) so that if a URI for a particular genus does not occur in Wikipedia it can be mapped to a URI.

I do a little editing of Wikipedia when I see things but I am not sure that that is the complete solution since I have had species pages I have added deleted later and the species moved to a genus page etc.

I can't make any promises until I figure out a way to automate this process better.

- Pete
--
------------------------------------------------------------------------------------
Pete DeVries
Department of Entomology
University of Wisconsin - Madison
445 Russell Laboratories
1630 Linden Drive
Madison, WI 53706
Email: pdev...@wisc.edu
TaxonConcept  &  GeoSpecies Knowledge Bases
A Semantic Web, Linked Open Data  Project
--------------------------------------------------------------------------------------
Reply all
Reply to author
Forward
0 new messages