Wikipedia:Articles for deletion/DBpedia

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was Withdrawn -- Collectonian (talk · contribs) 14:01, 16 November 2009 (UTC)[reply]

DBpedia[edit]

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

Unnotable project/organization. Fails WP:N, WP:ORG, and WP:WEB. Per lengthy talk page discussions, the bulk of the sources are articles written by those directly involved with DBpedia or Semantic Web itself. The few third-party, reliable sources either mention it in passing or are repeats of DBpedia's own press and soundbites. The Scientific America article is peer reviewed, but it was also written primarily by DBpedia members. This is true of the presentations and papers as well. They are reliable sources, of course, but also primary sources written by those directly involved and therefore cannot speak to its notability. Even the BBC link is just a glossary as part of its learning labs. Like the Amazon link, these are things submitted by those involved with DBpedia. The article itself was created by and is almost entirely written by DBpedia members, a clear conflict of issue. These are mostly self-identified or identified through the recent issue of the DBpedia templates being added to infobox documentation. Other than the single blog entry on the New York Times, noting that it was using the RDF which is used by both DBpedia and Freebase.

As a note, I strongly encourage anyone participating to actually read the talk page discussions, and to review the current references themselves, not just count them and say "its notable" as, again, almost all of them are from DBpedia itself. Note as well, I did also tag the article for possible merging to Semantic Web, however the continued disagreement between neutral editors and those involved in DBpedia and Semantic Web itself would seem to indicate that the notability issue needs to be discussed and dealt with in a more neutral, community based venue to determine if the article should exist at all. I would also ask that any involved in DBpedia who reply to this discussion please identify themselves, per Wikipiedia guidelines. -- Collectonian (talk · contribs) 17:32, 15 November 2009 (UTC)[reply]

Keep "most of them are from DBpedia themselves" versus exactly one person seems to think note is not established. OK, the article should not have been started autobiographically, but the suggestion of non-notability is simply incorrect. The statements made by Collectionian about peer review and citation, despite finding only dissent in talk, reveal a naivete which should be, by itself, sufficient for WP:SK. BarryNorton (talk) 17:50, 15 November 2009 (UTC)[reply]

Keep Not at all clear how it fails WP:N. Used, and mentioned by, independent, reliable organisations such as the BBC and New York Times. Most of these independent sources are blog posts, which is not the gold standard, but by not means irrelevant regarding notability. (As an aside, the Scientific American was not written by dbpedia contributors, as far as I can see). shellac (talk) 17:54, 15 November 2009 (UTC)[reply]

Keep. The significant BBC usage of Wikipedia data via DBpedia RDF is described in a paper that BBC staff co-authored with the DBPedia developers. Collectionian objects that "the bulk of the sources are articles written by those directly involved with DBpedia or Semantic Web itself", but provides no argument for why members of the (large, diverse and argumentative) Semantic Web community should be automatically presumed to be uncritical DBpedia fans. Many in the Semantic Web community would agree with the notability of DBpedia while questioning various specifics of the DBpedia approach. The talk page discussion also notes that the original publications about DBpedia by its developers have made significant impact on the computer science and information science literature, ie. they are highly cited by other researchers in the field. (Assuming we don't consider everyone working on Semantic Web themes intrinsically tainted...). A question: what kind of documentation might the BBC team be asked to provide, to help confirm the significance of DBpedia to their work? --DanBri (talk) 18:14, 15 November 2009 (UTC)[reply]

Significant to the BBC != notability. ~~It should also be noted that DanBri was a cowriter of one of the papers cited.~~ -- Collectonian (talk · contribs) 18:28, 15 November 2009 (UTC)[reply]

(I cowrote which paper? not aware I was self-citing anything). I am certainly working with other BBC folk (via NoTube project, www.notube.tv) who have made clear their understanding of the potential major impact of Wikipedia as classification system (in RDF via DBpedia), and I have discussed this at length in a recent seminar on subject classification, including critique of the idea that Wikipedia/DBpedia alone provides an adequate classification system for archival access. This is all on the public record. --DanBri (talk) 18:41, 15 November 2009 (UTC)[reply]
The BBC uses more than 5% of its 3.49 Billion GBP funding on Web projects (Television_licensing_in_the_United_Kingdom). Their opinion is notable. BarryNorton (talk) 18:38, 15 November 2009 (UTC)[reply]

Collectonian, could you please provide links for your claims? In this case, to which paper are you referring? And further, why is this relevant, are you claiming COI? —Preceding unsigned comment added by Pldms (talk • contribs) 12:45, November 15, 2009

Sorry, I mixed you up with Derivadow. With all the unsigned stuff on the talk page, things are getting confusing. -- Collectonian (talk · contribs) 18:55, 15 November 2009 (UTC)[reply]

No worries, Collectionian. But would you mind removing or annotating/updating your note here suggesting there's an issue with me co-authoring something? I did forget to sign some edits, sorry about that; and I'm not 100% familiar with the best conventions around here. --DanBri (talk) 19:06, 15 November 2009 (UTC)[reply]

Keep. The sources identified establish notability. The 2 sources where Sir Tim Berners-Lee uses dbpedia as his main example, if nothing else. (Notes: I have no connection to dbpedia or anything related to it, at all. I think Collectonian has done an ok (albeit less-than-friendly) job reminding editors of COI issues to be aware of, but is going beyond rational objections by suggesting merging or deleting this article) -- Quiddity (talk) 18:58, 15 November 2009 (UTC)[reply]
Keep. (disclosure: I work for the BBC and coauthored one of the papers referenced - however I did not add the reference to the dbpedia page; I did add the link to Tim Berners-Lee describing dbpedia as "one of the more famous parts of the Linked Data Project."). Dbpedia has been the subject of numerous peer reviewed papers (as referenced), it has been talked about on a number of occasions by the inventor of the web and is the subject of the paper I co-authored not because I work for dbpedia and wish to promote it but because it is an important component in the web architecture for significant chunks of bbc.co.uk. —Preceding unsigned comment added by Derivadow (talk • contribs) 19:08, 15 November 2009 (UTC)[reply]
Keep. No need to add to the above, but FWIW http://scholar.google.com/scholar?q=dbpedia&hl=en&btnG=Search obtains 856 'hits'. Mike Linksvayer (talk) 19:53, 15 November 2009 (UTC)[reply]

Google hits are not signs of notability, particular when the bulk of the papers are, again, written by DBpedia members itself. -- Collectonian (talk · contribs) 20:02, 15 November 2009 (UTC)[reply]

These aren't general Google hits, but Google scholar; generally academic research papers. As the Talk page mentioned citeseer 2007 ranks the 'What have Innsbruck and Leipzig in common? Extracting Semantics from Wiki Content' paper very highly. Collectionian, can you clarify what you mean by 'DBpedia members'? Are you suggesting that amongst the 856 hits on Google Scholar, there's nothing much of interest except by the people who actually created DBpedia?--DanBri (talk) 20:12, 15 November 2009 (UTC)[reply]

Can you point to any that are not written by those involved with the project? -- Collectonian (talk · contribs) 20:17, 15 November 2009 (UTC)[reply]

Sure, just do a search for "dbpedia -bizer -auer -kobilarov -cyganiak -erling -idehen -hellmann -Jentzsch -Kreis -Lehmann -Schüppel" (where those are the names listed on the dbpedia.org credits page); it still matches 333 results. Doubtless some of the other authors are DBpedia enthusiasts, since they've been moved to write on the topic. And since the Semantic Web community try to foster collaboration, there's a good chance they've even tried to *help* the DBpedia effort too. Probably you could exclude a few more hits from the list by excluding practical-minded, collaborative and helpful people who're on the record as supporting the DBpedia effort. But really, why bother? --DanBri (talk) 20:27, 15 November 2009 (UTC)[reply]
"Google hits" is an extremely disingenuous term to use here. Plain Google hits are keyword-based, Google Scholar hits are much more sophisticated. Given that you remain the only one suggesting deletion I think you'd take a bit more care than making false accusations and adopting misleading terminology. (Dan just beat me to this) BarryNorton (talk) 20:13, 15 November 2009 (UTC)[reply]

Google scholar hits are not that much more sophisticated, and do not include only peer review works, but also student thesis and other sources. They also do not filter out those written by those involved with DBpedia. -- Collectonian (talk · contribs) 20:17, 15 November 2009 (UTC)[reply]

"Google scholar hits are not that much more sophisticated" - I'm sorry, but do you even understand what keyword-based search is? You seem completely out of your depth here. BarryNorton (talk) 20:20, 15 November 2009 (UTC)[reply]

Google scholar hits are clearly no more sophisticated or different from regular google results. The only difference is what it searches. As noted below, it is NOT limited to academic and reliable sources only, any more than any Google search result is. -- Collectonian (talk · contribs) 21:24, 15 November 2009 (UTC)[reply]
This stands. If an independent editor feels it breaks policy then I'm prepared to take the consequences. I will not be censored to further your vendetta. Furthermore unless you find one vote to support your proposed deletion, or undo it, I will seek mediation tomorrow. BarryNorton (talk) 20:32, 15 November 2009 (UTC)[reply]

Keep - 800+ Google Scholar hits, and the main paper has over 200 citations. No question that this has had significant influence and is notable. --Stephan Schulz (talk) 21:09, 15 November 2009 (UTC)[reply]

Again, Google Scholar hits do not show notability. Google scholar results are NOT limited to only peer review works. In fact, my own junior college papers and personal essays which are purely self-published on my personal website appear in Google Scholar search results. Again, please show specific papers showing notability. -- Collectonian (talk · contribs) 21:24, 15 November 2009 (UTC)[reply]

I've made my point sufficiently and stand by it. --Stephan Schulz (talk) 21:35, 15 November 2009 (UTC)[reply]

Strong, speedy keep - Apart from the 333 third party Google Scholar hits reported above, it reports 68 Google books hits, which seem to be from mostly third-party authors. Even if this was not the case (but it is), being the subject of tons of peer-reviewed publications and on the media surely makes it notable, even if most sources were from involved authors. But no need to worry: there are lots of third party RS here. --Cyclopia ^talk 21:28, 15 November 2009 (UTC)[reply]
Keep. It seems pointless to try to add more to the overwhelming evidence of notability above – all of it, note, added by people who are not directly involved with the DBpedia project (being instead users of it as a core Semantic Web technology). Collectonian appears to believe that anyone who as much as knows about the DBpedia project is ipso facto no longer neutral enough to be allowed an opinion (I have literally never heard of any of the articles that Collectonian highlights on her talk page, apart from 101 Dalmations, but that doesn't mean they're non-notable) NormanGray (talk) 21:29, 15 November 2009 (UTC)[reply]
Keep. The number of citations of papers about DBpedia show its notability. Nloth (talk) 21:54, 15 November 2009 (UTC) Edit: Disclaimer - I have no involvement with Semantic Web projects, and I would object VERY strongly to being characterized as a SemWeb fan. However, I once did have a paper that had "Linked Data" in the title presented at a conference.[reply]
Keep Passes WP:RS and WP:WEB. Warrah (talk) 23:13, 15 November 2009 (UTC)[reply]
Keep The article is written well and passes notability. - 4twenty42o (talk) 23:45, 15 November 2009 (UTC)[reply]
Keep DBpedia is widely covered in published literature. For instance, it is described in this book http://my.safaribooksonline.com/9780596802141 and used in the examples. I don't believe DBpedia project members wrote those 68 books listed by Google book search. It seems notable enough to be covered in class. http://web.cecs.pdx.edu/~mm/nonstandard-computation2008/index.html I have no involvement in this subject apart from being a computer science student with an interest in future information lookup methods. Leif Warner Abimelech (talk) 01:22, 16 November 2009 (UTC)[reply]
Keep for all the reasons listed above. Several peer reviewed publications, conference proceedings, and the heavily referenced LNCS chapter are listed in the Google Scholar search listed. Several books mentioned above reference DBPedia. Several peer reviewed papers reference the use of DBPedia in methodology. Collectively, I think this clearly demonstrates notability. I have reviewed Talk:DBpedia as suggested by the nom, but I fail to see how it, or the nomination statement demonstrate that this is lacking in notability. FWIW I have had no involvement in DBpedia and hadn't heard of it until minutes ago -- Samir 08:41, 16 November 2009 (UTC)[reply]
Comment this article is in desperate need of a cleanup. But right now the article accomplishes little else but hammering home how notable it is. 90% of the article is a list of numbers with no context, mentions of other "interlinked" datasets with no mention of who is involved (an uneducated reader might think the CIA and US Census Bureau are using the DBpedia data), plus TBL's comment which doesn't really tell me anything (is DBpedia famous the same way TBL is famous?). Since you guys claim to know something about DBpedia, can you improve the article by answering some of these questions?

Who started the DBpedia project, and who maintains it?
How often is the dataset updated, and who does it?
What is the process through which the dataset is built (algorithms & software used, etc.)
How is DBpedia actually used? (Beyond "NYT includes links" and "BBC uses it to organize stuff".. I have no idea what that even means!)
In fact, how is OpenCalais even based on the NYT? I scanned the references and don't see any connection.--Jonovision (talk) 11:14, 16 November 2009 (UTC)[reply]

Jonovision, you're right it certainly has room for improvement. I tried recently to add an example, which shows - albeit for a technical audience who understand the idea of a query language - what DBpedia is capable of. My edit was reverted though ('not a howto'). The threat of deletion seems to have led to an overfocus on its notability at the expense of other information. What's the process for moving beyond the 'articles for deletion' stage, so that (if it's a keeper...) those who have the answers to these questions might feel it worthwhile contributing again? —Preceding unsigned comment added by DanBri (talk • contribs) 11:27, 16 November 2009 (UTC) oops sorry always forget signing needed here! --DanBri (talk) 11:41, 16 November 2009 (UTC)[reply]

Thanks for pointing me to your edit. I guess your detailed example was a bit longish for this article (since it ends up focusing more on SPARQL than DBpedia itself), but as someone who doesn't know anything about DBpedia, I found these bits particularly useful: "By extracting factual information from thousands of Wikipedia page, DBpedia makes it possible to find the answers to questions where the relevant information is spread across several different Wikipedia entries. Since DBpedia normalises information into a single database, the following query can be asked without needing to know exactly which entry carries each fragment of information." Maybe that can be worked back into the article. --Jonovision (talk) 11:46, 16 November 2009 (UTC)[reply]

FWIW, I fully agree with Jonovision. --Cyclopia ^talk 12:37, 16 November 2009 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.