- From: Paul Groth <p.t.groth@vu.nl>
- Date: Fri, 20 Apr 2012 08:55:00 +0200
- To: Daniel Garijo <dgarijo@delicias.dia.fi.upm.es>
- Cc: Timothy Lebo <lebot@rpi.edu>, Luc Moreau <L.Moreau@ecs.soton.ac.uk>, "public-prov-wg@w3.org" <public-prov-wg@w3.org>
Hi Daniel, All, Sorry to be late jumping in on this thread. Definitely, wasQuotedFrom is modeling what it is intended to model now. Tim's examples are good. The main case I'm trying to support is <blockquote> on the web. This happens all the time in blogs. They quote from a newspaper and then add some commentary. You often see many blogs (e.g. [1]) that look take a piece of content and reuse it from another site. You want to identify that content as an entity and link it back to the source ([2]) :blockquoteX prov:wasQuotedFrom :newspaperArticleY Another way to read it is blockquoteX is a quote from newspaperArticleY but we don't do that because we put everything in the past tense. So I would argue for it to stand as is. Does that make sense? Paul [1] http://www.washingtonpost.com/blogs/ezra-klein/post/want-lower-tax-rates-hire-a-lobbyist/2012/04/18/gIQA8X3hQT_blog.html [2] Note, it's a pain to mine this information from site because the blockquote is often not directly after the link that tells you the provenance that's why we need some structured data. On Thu, Apr 19, 2012 at 11:44 PM, Daniel Garijo <dgarijo@delicias.dia.fi.upm.es> wrote: > Hi Tim, > I see your point. IMO, if you wanted to separate both the quotation from the > rest of the > post you could still do it creating a separate entity. Right now we are > forcing the user > either to do so or to use another relationship. > > Thanks for the examples. I still find a bit funny that I can use > wasQuotedFrom for copying and downloading files > and I can't use it for saying that my post actually quoted another > post/article. I have just > realized that my expanded terms example is not completely right, so I'll > have to change it. > > Since both of my suggestions have been droped, I don't have a better name > for the moment. > I'll try to think of another one, and if I don't manage to come up with a > new one I'll close the issue on monday. > > However, I would still like to know Paul's point of view on this thread. His > "5 simple provenance statements" > example was about posts (although no reference to wasQuotedFrom was there). > > > Thanks, > Daniel > > > 2012/4/19 Timothy Lebo <lebot@rpi.edu> >> >> Daniel, >> >> On Apr 19, 2012, at 4:53 PM, Daniel Garijo wrote: >> >> Hi Tim, Luc. >> From what I understood, I thought that wasQuotedFrom was way less >> restrictive. >> For instance, if a blogger writes an opinion and quotes another article in >> a blog post >> I would expect him to assert that the post wasQuotedFrom the article: >> >> :post prov:wasQuotedFrom :article >> (Therefore the prov:hadQuoteFrom would make sense, as in your example) >> >> >> Instead, if I understood correctly, we are forcing him to create an >> intermediate entity just for the quote >> that is used in the publication activity which generated the article. >> >> >> Yes. >> >> I think we see each type of modeling (the "pedantic via direct quote way" >> and the "abbreviated post-to-post way") >> >> I'm glad that it is clear, so that the WG can decide on which they want. >> >> >> I can't see how that is scruffy provenance >> (wasn't it supposed to be a shortcut??): >> >> :quote a prov:Entity; >> prov:wasQuotedFrom :article. >> >> :publActivity a prov:Activity; >> prov:used :quote; >> prov:generated :post. >> >> >> You can omit the activity and use a derivedFrom like I did in my example. >> >> :post prov:wasDerivedFrom :quote . >> >> >> :post a prov:Entity; >> prov:wasGeneratedBy :publActivity. >> >> Since it was a kind of derivation, I assumed that if you added additional >> stuff to the entity that is repeating >> some of all of the other entity it would be a quotation… >> >> >> If we define it like this, how do we distinguish which part of the entity >> is quoted and which part of the entity is original? >> (pedantic, proper hat is clearly on here) >> >> Appart from the notion of retweeting, then I don't find the shortcut very >> useful, to tell you the truth. >> >> >> If you don't want the granularity, then use wasQuotedFrom's super >> properties: wasDerivedFrom or tracedTo. They give you the abstraction you >> want, without the details you aren't concerned about. >> >> >> Downloading a file is a very common wasQuotedFrom. >> >> :myFile >> a foaf:Document, prov:Entity; >> prov:atLocation <file:///Users/me/files/working.html>; >> prov:wasQuotedFrom <http://www.w3.org/2011/prov/wiki/WorkingDrafts>; >> :size "45"^^:kilobytes; >> . >> >> >> >> >> Any sort of copy-paste operation is naturally modeled with wasQuotedFrom: >> >> :copy >> a prov:Activity; >> prov:wasAssociatedWith :tlebo; >> prov:generated :clipboard_contents; >> . >> >> :tlebo >> a foaf:Account; >> prov:atLocation :tim_laptop; >> . >> >> :clipboard_contents >> a prov:Entity; >> prov:value "Provenance Data Model (PROV-DM)"; >> prov:wasQuotedFrom :page; >> prov:wasInvalidatedBy :the_next_copy_operation; >> . >> >> :page >> a prov:Entity; >> dcterms;date "2012-04-13"; >> prov:specializationOf >> <http://www.w3.org/2011/prov/wiki/WorkingDrafts>; >> . >> >> :tech_report >> a prov:Entity, :TechReport; >> prov:used :clipboard_contents; >> prov:wasAttributedTo <http://data.semanticweb.org/person/timothy-lebo>; >> . >> >> >> -Tim >> >> >> People >> use to comment what they are quoting, IMO. >> >> Thanks, >> Daniel >> >> >> 2012/4/19 Timothy Lebo <lebot@rpi.edu> >>> >>> >>> On Apr 19, 2012, at 3:31 PM, Daniel Garijo wrote: >>> >>> Hi Luc, >>> hmmm and what about my other suggestion, "hadQuoteFrom" ? >>> >>> >>> Daniel, >>> >>> I'm not in favor of changing it. >>> >>> I think your suggestion of hadQuoteFrom changes the meaning of the >>> definition, where the quote is not THE thing taken from the original source, >>> but CONTAINS something taken from the original source (and thus a subsequent >>> derivation). >>> >>> e.g. >>> >>> :composite_tweet >>> a :Tweet; >>> prov:value "I have always loved the #blah. Like @Abe said, "Four score >>> and seven years ago"; >>> daniel:hadQuoteFrom <http://dbpedia.org/resource/Gettysburg_Address>; >>> # This is not the meaning of the current definition "the repeat of (some >>> or all of) an entity.." >>> prov:wasAttributedTo twitter:timrdf, >>> >>> <http://dbpedia.org/resource/Abraham_Lincoln>; >>> prov:qualifiedAttribution [ >>> a prov:Attribution; >>> prov:agent <http://dbpedia.org/resource/Abraham_Lincoln>; >>> prov:hadRole "contributor", "quoted"; >>> ] >>> prov:qualifiedAttribution [ >>> a prov:Attribution; >>> prov:agent twitter:timrdf; >>> prov:hadRole "composer", "quoter"; >>> ] >>> prov:wasDerivedFrom :actual_phrase; ## This derivation shows the >>> distinction between the meaning of what you propose and how it is currently >>> defined. >>> ] >>> >>> is NOT the same as >>> >>> :actual_phrase >>> a :Phrase; >>> prov:value "Four score and seven years ago"; >>> prov:wasQuotedFrom <http://dbpedia.org/resource/Gettysburg_Address>; >>> prov:wasAttributedTo twitter:timrdf; >>> . >>> >>> <http://dbpedia.org/resource/Gettysburg_Address> >>> a frbr:Work; >>> prov:wasAttributedTo <http://dbpedia.org/resource/Abraham_Lincoln>; >>> . >>> >>> <http://en.wikipedia.org/wiki/Gettysburg_Address> >>> a foaf:Document; >>> prov:specializationOf >>> <http://dbpedia.org/resource/Gettysburg_Address>; >>> . >>> >>> >>> -Tim >>> >>> >>> >>> >>> >>> Thanks, >>> Daniel >>> >>> 2012/4/19 Luc Moreau <L.Moreau@ecs.soton.ac.uk> >>>> >>>> Daniel, >>>> We started with wasQuoteOf >>>> http://www.w3.org/TR/2011/WD-prov-dm-20111018/#quotation >>>> But moved away because not clear. >>>> >>>> Luc >>>> >>>> >>>> Professor Luc Moreau >>>> Electronics and Computer Science >>>> University of Southampton >>>> Southampton SO17 1BJ >>>> United Kingdom >>>> >>>> On 19 Apr 2012, at 17:39, "Daniel Garijo" >>>> <dgarijo@delicias.dia.fi.upm.es> wrote: >>>> >>>> Hi Luc, >>>> the definition on DM is very clear to me. >>>> >>>> What makes me feel a bit unconfortable is that while I can understand >>>> what is on the domain >>>> and what is on the range on each of the other properties, for this one I >>>> think it is a bit confusing. >>>> (When I say domain and range, I refer to what is being quoted (original) >>>> and what is the quote). >>>> >>>> I have asked 3 colleagues in my lab to tell me what did they think they >>>> were the range and the domain >>>> of the property with an example, (without looking at the definition of >>>> the DM). One of them agreed with the DM, >>>> another one guessed wrong and the last one encouraged me to change the >>>> naming because "it made >>>> no much sense" to him. >>>> >>>> I'm not sure if users that assert scruffy provenance will come to the DM >>>> to read all the definitions, >>>> and that is why to make sure this kind of things are very clear for >>>> everyone. Thus, I don't propose >>>> to change the definitions, I just suggest to rename "wasQuotedFrom" to >>>> either: >>>> "wasQuoteOf" or "hadQuoteFrom". >>>> >>>> Thanks, >>>> Daniel >>>> >>>> 2012/4/19 Luc Moreau <L.Moreau@ecs.soton.ac.uk> >>>>> >>>>> Hi Daniel, >>>>> >>>>> This is the current definition of quotation. Is there still a concern >>>>> with it? >>>>> >>>>> Cheers, >>>>> Luc >>>>> >>>>> 4.3.3 Quotation >>>>> >>>>> A quotation is the repeat of (some or all of) an entity, such as text >>>>> or image, by someone other than its original author. >>>>> >>>>> Quotation is a particular case of derivation in which entity e2 is >>>>> derived from an original entity e1 by copying, or "quoting", some or all of >>>>> it. A quotation relation, written wasQuotedFrom(id,e2,e1,ag2,ag1,attrs) in >>>>> PROV-N, has: >>>>> >>>>> id: an optional identifier for the relation; >>>>> quote: an identifier (e2) for the entity that represents the quote (the >>>>> partial copy); >>>>> original: an identifier (e1) for the original entity being quoted; >>>>> quoterAgent: an optional identifier (ag2) for the agent who performs >>>>> the quote; >>>>> originalAgent: an optional identifier (ag1) for the agent to whom the >>>>> original entity is attributed; >>>>> attributes: an optional set (attrs) of attribute-value pairs >>>>> representing additional information about this relation. >>>>> >>>>> >>>>> >>>>> On 04/19/2012 11:28 AM, Provenance Working Group Issue Tracker wrote: >>>>>> >>>>>> PROV-ISSUE-352 (rename-WasQuotedFrom): A better name for wasQuotedFrom >>>>>> [prov-dm] >>>>>> >>>>>> http://www.w3.org/2011/prov/track/issues/352 >>>>>> >>>>>> Raised by: Daniel Garijo >>>>>> On product: prov-dm >>>>>> >>>>>> Currently, the DM says: >>>>>> A quotation record, written wasQuotedFrom(e2,e1,ag2,ag1,attrs) in >>>>>> PROV-ASN, contains: >>>>>> quote: an identifier e2, identifying an entity record that >>>>>> represents the quote; >>>>>> quoted: an identifier e1, identifying an entity record >>>>>> representing what is being quoted; >>>>>> ... >>>>>> >>>>>> However, if we say that e2 wasQuotedFrom e1 it may look like entity e1 >>>>>> is the one quoting e2 (since we are saying that e2 was quoted). >>>>>> >>>>>> I think it would be more clear if we rename the property with e2 >>>>>> wasQuoteOf e1, or e2 hadQuoteFrom e1. >>>>>> >>>>>> Thoughts? >>>>>> Thanks, >>>>>> Daniel >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> Professor Luc Moreau >>>>> Electronics and Computer Science tel: +44 23 8059 4487 >>>>> University of Southampton fax: +44 23 8059 2865 >>>>> Southampton SO17 1BJ email: l.moreau@ecs.soton.ac.uk >>>>> United Kingdom http://www.ecs.soton.ac.uk/~lavm >>>>> >>>>> >>>> >>> >>> >> >> > -- -- Dr. Paul Groth (p.t.groth@vu.nl) http://www.few.vu.nl/~pgroth/ Assistant Professor Knowledge Representation & Reasoning Group Artificial Intelligence Section Department of Computer Science VU University Amsterdam
Received on Friday, 20 April 2012 06:55:30 UTC