Hello, The issue is hard to explain, as there are still quite some bugs on dbpedia-live, which we only found after letting it run for a while. We optimized the speed and will soon reload DBpedia 3.4 and also load all changes since September, which will fix the missing or double abstracts. Here is how it should work: Every page has static abstracts (the ones you know). These abstracts are the same as as in 3.4 and will remain static. abstract_live is the abstract extracted for the last revision. The main reason why there are two is, that there are two different responsible extractors: - A slow one, which produces abstracts with better quality and produced the abstracts for 3.4. - A fast one (factor 10-100) , which produces abstracts with more or less acceptable quality and produces comment_live and abstract_live Once we improve the speed of the "better" AbstractExtractor, they will be merged again, but this could take quite some time as it is a complicated and expensive process( It involves parsing Wiki syntax, extending a MediaWiki and synchronizing template definitions...)
Perhaps, it will even stay like that and we will only refresh the static abstract information with each available Wikipedia dump. So there would always be a live english (for now) version and a static one with better quality and for all languages. I think, it might not be the worst solution. Regards, Sebastian Stephen Betts schrieb: > We (in the BBC Search team) are still seeing the problem below. > > To recap, Dave wrote: > > I'm having difficulty querying dbpedia using the sparql interface on > > http://dbpedia-live.openlinksw.com/sparql > > I'm requesting the label, abstract and a list of redirects, and > approx. 20% of the time I'm just getting the redirects. > > > For example using a simple SPARQL query like > > SELECT ?abstract WHERE { > <http://dbpedia.org/resource/Ferrari> > <http://dbpedia.org/property/abstract> ?abstract . > FILTER ( langMatches( lang(?abstract), 'en') || ! > langMatches(lang(?abstract),'*') ) > } > > on dbepdia.org/sparql gives the abstract that you’d expect; running it > on dbpedia-live.openlinksw.com/sparql doesn’t find it. However if you > use abstract_live rather than abstract then it does find something – > although surprisingly it finds two different abstracts! > > > Just to be explicit, running this query > > SELECT ?abstract WHERE { > <http://dbpedia.org/resource/Ferrari> > <http://dbpedia.org/property/abstract_live> ?abstract . > FILTER ( langMatches( lang(?abstract), 'en') || ! > langMatches(lang(?abstract),'*') ) > } > > on dbpedia-live.openlinksw.com/sparql does give a result, in fact it > gives two! > > So really I have three main questions (although any information would > be useful): > > 1. What is the http://dbpedia.org/property/abstract_live property? > 2. Cam we safely use it instead of > http://dbpedia.org/property/abstract > 3. Why does it have more than one value? > > > We are seeing this on around 20% of the resources that we are > interested in, and is preventing us from using the Live DBPedia at > present. > > Thanks very much, > > Stephn. > > > On 15/1/10 12:06, "[email protected]" > <[email protected]> wrote: > > > ----------------------------- > > Message: 2 > Date: Fri, 15 Jan 2010 11:05:54 +0000 > From: David Spacey <[email protected]> > Subject: [Dbpedia-discussion] Inconsistent results from sparql queries > To: [email protected] > Message-ID: <[email protected]> > Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed > > Hi, > > I'm having difficulty querying dbpedia using the sparql interface on > > http://dbpedia-live.openlinksw.com/sparql > > I'm requesting the label, abstract and a list of redirects, and > approx. 20% of the time I'm just getting the redirects. With an > internal instance, running an older snapshot of dbpedia, I'm seeing > the same problem for around 1% of queries. In every example I've > checked, the dbpedia id is still current. > > So, here are some examples. > > Failing on dbpedia-live only > Adolf_Hitler > Chicago > Dora_Bryan > Emma_Thompson > Ferrari > > Failing on snapshot only > Austriamicrosystems > Valery_Gergiev > Lady_Rachel_Billington > > Failing on both > G20 > Darcy_Edwards > > This is the query I'm using, where '$lod_id' is one of the strings > listed above. > > SELECT ?label, ?abstract, ?redirect, COUNT(?wikilink) > WHERE { > { > <http://dbpedia.org/resource/$lod_id> <http://dbpedia.org/ > property/abstract> ?abstract . > FILTER ( langMatches( lang(?abstract), 'en') || ! langMatches > (lang(?abstract),'*') ) . > <http://dbpedia.org/resource/$lod_id> <http://www.w3.org/2000/01/ > rdf-schema#label> ?label . > FILTER ( langMatches( lang(?label), 'en') || ! langMatches(lang(? > label),'*') ) . > } > UNION > { > OPTIONAL { > ?redirect <http://dbpedia.org/property/redirect> <http:// > dbpedia.org/resource/$lod_id> . > OPTIONAL { ?wikilink <http://dbpedia.org/property/wikilink> ? > redirect } > } > } > } > > Can anyone suggest why this might fail for valid entries? > > TIA > > Dave Spacey > > > http://www.bbc.co.uk/ > This e-mail (and any attachments) is confidential and may contain > personal views which are not the views of the BBC unless > specifically stated. > If you have received it in error, please delete it from your system. > Do not use, copy or disclose the information in any way nor act in > reliance on it and notify the sender immediately. > Please note that the BBC monitors e-mails sent or received. > Further communication will signify your consent to this. > > > > > ------------------------------ > > > ------------------------------------------------------------------------------ > Throughout its 18-year history, RSA Conference consistently > attracts the > world's best and brightest in the field, creating opportunities > for Conference > attendees to learn about information security's most important > issues through > interactions with peers, luminaries and emerging and established > companies. > http://p.sf.net/sfu/rsaconf-dev2dev > > ------------------------------ > > _______________________________________________ > Dbpedia-discussion mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion > > > End of Dbpedia-discussion Digest, Vol 35, Issue 11 > ************************************************** > > > > http://www.bbc.co.uk > This e-mail (and any attachments) is confidential and may contain > personal views which are not the views of the BBC unless specifically > stated. > If you have received it in error, please delete it from your system. > Do not use, copy or disclose the information in any way nor act in > reliance on it and notify the sender immediately. > Please note that the BBC monitors e-mails sent or received. > Further communication will signify your consent to this. > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------------ > Download Intel® Parallel Studio Eval > Try the new software tools for yourself. Speed compiling, find bugs > proactively, and fine-tune applications for parallel performance. > See why Intel Parallel Studio got high marks during beta. > http://p.sf.net/sfu/intel-sw-dev > ------------------------------------------------------------------------ > > _______________________________________________ > Dbpedia-discussion mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion > -- Dipl. Inf. Sebastian Hellmann Department of Computer Science, University of Leipzig Homepage: http://bis.informatik.uni-leipzig.de/SebastianHellmann Research Group: http://aksw.org ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Dbpedia-discussion mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
