[Data-modeling] Add page number(s) to "Quotation"

Ed Laurent spatial.db at gmail.com
Thu Dec 4 03:04:10 UTC 2008


I modeled "Cited quotation" in my LitCentral base (
http://litcentral.freebase.com/) that has a property of "Quotation" (set as
the common "Quotation" type) with disambiguation for "Page number" (machine
readable text). For the time being, I also added a "Relevance notes" (short
text) property as disambigution. I may modify that later, depending on use.
This form of "Cited quotation" model adds denormalization to quotations
(duplicate "Source" properties) but I don't see a way around that.

I originally modeled "Quotation and page number", with a "Quotation"
property delegated from the "Quotation" common type. However, it didn't seem
right to orphan quotations from their sources - a visible reciprocating link
cannot be established between source and quotation using a CVT (maybe I'm
wrong?).

Another property I added to "Cited quotation" is "Freebase cross reference",
which is a reciprocating link between topics that are not properties of each
other. It seemed a good way to collect relevance data that might be helpful
for informing the modeling of future properties.

I'm including the "Cited quotation" type in "Ecological publication" of my
Bird Conservation base (http://birdconservation.freebase.com) along with the
common "Journal article" type (includes "Written work" and "Scholarly work")
because that will be my primary use for it. Instances of ecological
publications that are not journal articles (e.g., technical reports) will
need to have that type removed by hand. Regardless, its Included types will
often be appropriate.

Currently, "Full view" must be used to view the LitCentral types included
within "Ecological document" but I'm hopeful that there will be an option to
display all Included types, regardless of whether they are common types or
not, in an upcoming release.

I'm open to better ways to do this. Thanks for everyone's input!

-Ed


On Wed, Dec 3, 2008 at 8:05 PM, Ed Laurent <spatial.db at gmail.com> wrote:

> <punt>
>
>
>
> On Wed, Dec 3, 2008 at 7:54 PM, Jeff Prucher <jeff at metaweb.com> wrote:
>
>>  If all you need to cite are articles, technical reports, and the like,
>> I'd suggest punting the book edition problem entirely for this use case.
>>
>> Jeff
>>
>>  ------------------------------
>> *From:* data-modeling-bounces at freebase.com [mailto:
>> data-modeling-bounces at freebase.com] *On Behalf Of *Ed Laurent
>> *Sent:* Wednesday, December 03, 2008 4:26 PM
>>
>> *To:* Freebase data modeling mailing list
>> *Subject:* Re: [Data-modeling] Add page number(s) to "Quotation"
>>
>> Great suggestions. However, linking to book edition is not an option for
>> journal articles. In fact, "Book edition" doesn't even include "Written
>> work". This has been a thorny issue for me for other reasons too and I'm
>> slowly understanding why "Book edition" is modeled this way but it seems to
>> make properties such as "Cited quotation" impossible(?) to apply to all
>> written works.
>>
>> Maybe a generic "Cited quotation" type could be implemented as a property
>> of all written works if it had separate "Written work cited" and "Book
>> edition cited" properties... ?
>>
>> -Ed
>>
>>
>> On Wed, Dec 3, 2008 at 7:06 PM, Tadhg O'Higgins <tadhg at metaweb.com>wrote:
>>
>>> +1 for this suggestion of Jeff's.
>>>
>>> Perhaps a Cited Quotation type, with all properties delegated from
>>> Quotation other than a References property?
>>>
>>> Actually, it's possible that the best way to do this would be to create
>>> the Citation type that Tim suggested, have that take Book Edition (plus
>>> page number, etc.), and then give Cited Quotation type a property that
>>> accepted multiple Citation instances, so that e.g "There is no calamity
>>> greater than lightly engaging in war." would have all the properties it
>>> does here:
>>> http://www.freebase.com/view/guid/9202a8c04000641f800000000519187c
>>> plus additional links to a CVT pointing to
>>> http://www.freebase.com/view/guid/9202a8c04000641f8000000009a0b66b and a
>>> page number.
>>>
>>> Tadhg
>>>
>>> Jeff Prucher wrote:
>>> > You could create your own type, but delegate the relevant properties
>>> > in from the quotation commons type.  This would feed your quotes into
>>> > the quotations commons, but allow you to add properties specific to
>>> > your needs.
>>> >
>>> > Jeff
>>> >
>>> >
>>> ------------------------------------------------------------------------
>>> >     *From:* data-modeling-bounces at freebase.com
>>> >     [mailto:data-modeling-bounces at freebase.com] *On Behalf Of *Ed
>>> Laurent
>>> >     *Sent:* Wednesday, December 03, 2008 3:19 PM
>>> >     *To:* Freebase data modeling mailing list
>>> >     *Subject:* Re: [Data-modeling] Add page number(s) to "Quotation"
>>> >
>>> >     Nope, I actually want to store quotes from published sources. I've
>>> >     talked with several editors who saw no legal issues with quoting a
>>> >     couple sentences (let me know if this is not the case and I'll
>>> >     stop). I want to do this because the quote gives more context to a
>>> >     bibliographic reference than the citation. It is also useful when
>>> >     a publication does not have an abstract.
>>> >
>>> >     -Ed
>>> >
>>> >
>>> >     On Wed, Dec 3, 2008 at 5:52 PM, Tim Kientzle <tim at metaweb.com
>>>  >     <mailto:tim at metaweb.com>> wrote:
>>> >
>>> >         Ed,
>>> >
>>> >         Sounds like what you're looking for might be better described
>>> as a
>>> >         "citation"
>>> >         or a "bibliographic reference."  Agree that there's overlap
>>> >         between
>>> >         that and
>>> >         the informal use of the term "quotation," but given the
>>> different
>>> >         usages, it
>>> >         may make sense to distinguish them.
>>> >
>>> >         Tim Kientzle
>>> >
>>> >         On Dec 3, 2008, at 2:46 PM, Ed Laurent wrote:
>>> >
>>> >         > Um, I'm interpreting your responses as that I should create
>>> >         my own
>>> >         > quotation type. Most of the sources I want to quote (e.g.,
>>> >         journal
>>> >         > articles, technical reports) are only published once.
>>> >         >
>>> >         > I agree that page number isn't appropriate for the cases you
>>> >         suggest
>>> >         > but it is for mine and I try to use common types whenever
>>> >         possible.
>>> >         > There have been other situations where I've argued the other
>>> >         side of
>>> >         > properties needing to apply to all instances of types so this
>>> >         > conversation has taken a somewhat amusing turn.
>>> >         >
>>> >         > For now I guess I'll just model my own quotation type and try
>>> to
>>> >         > keep it similar to the common one in case there are
>>> >         suggestions for
>>> >         > merging them later.
>>> >         >
>>> >         > Thanks!
>>> >         > -Ed
>>> >         >
>>> >         >
>>> >         > On Wed, Dec 3, 2008 at 5:34 PM, Jeff Prucher
>>>  >         <jeff at metaweb.com <mailto:jeff at metaweb.com>> wrote:
>>> >         > I agree -- page number isn't really applicable here.
>>> >         Technically,
>>> >         > since we
>>> >         > do have distinct types for book editions, it would be
>>> >         possible to
>>> >         > link the
>>> >         > source to a book edition, rather than to a book, but this
>>> >         seems like
>>> >         > a bad
>>> >         > idea. For one, it would be very hard to simply find all the
>>> >         quotes
>>> >         > from a
>>> >         > single book, since they might be spread across multiple
>>> >         editions.
>>> >         > But for
>>> >         > non-book works (plays, poems, essays, articles, etc.), a
>>> >         page number
>>> >         > would
>>> >         > be completely meaningless, since we do not have first-class
>>> >         topics
>>> >         > for every
>>> >         > appearance of these types of works.
>>> >         >
>>> >         > The reason we used "author" as the name of the property is
>>> >         because
>>> >         > we're
>>> >         > looking for the person who authored what was said (whether
>>> >         it was
>>> >         > written or
>>> >         > spoken impromptu), rather than the person who said it.
>>> >          Although, it
>>> >         > must be
>>> >         > said, this is still a problem with movie and TV quotes, so
>>> >         maybe it
>>> >         > doesn't
>>> >         > really make a difference in that regard.
>>> >         >
>>> >         > Jeff
>>> >         >
>>> >         > > -----Original Message-----
>>> >         > > From: data-modeling-bounces at freebase.com
>>> >         <mailto:data-modeling-bounces at freebase.com>
>>> >         > > [mailto:data-modeling-bounces at freebase.com
>>>  >         <mailto:data-modeling-bounces at freebase.com>] On Behalf Of
>>> >         > > Kirrily Robert
>>> >         > > Sent: Wednesday, December 03, 2008 1:54 PM
>>> >         > > To: Freebase data modeling mailing list
>>> >         > > Subject: Re: [Data-modeling] Add page number(s) to
>>> "Quotation"
>>> >         > >
>>> >         > > Wouldn't page number be dependent on edition?  On what page
>>> >         > > does Alice's "Curiouser and curiouser!" appear, for
>>> example?
>>> >         > >
>>> >         > > K.
>>> >         > >
>>> >         > > On Dec 3, 2008, at 1:15 PM, Ed Laurent wrote:
>>> >         > >
>>> >         > > > I'm thinking of using the "Quotation" type to store
>>> >         quotes from
>>> >         > > > published works but need a page number property because
>>> >         > > page numbers
>>> >         > > > are often required when citing quotations. It appears
>>> >         that the
>>> >         > page
>>> >         > > > number of the quote would need to be added as a
>>> >         > > disabiguator so that
>>> >         > > > this data can be entered when the "Quotation" type is
>>> >         linked as a
>>> >         > > > property of another type. Does this seem reasonable?
>>> >         > > >
>>> >         > > > I also see that "Author" (linked to "Quotations" property
>>> of
>>> >         > > > "Person") is a property of "Quotation", and I'm assuming
>>> >         > > that is for
>>> >         > > > instances when the quotation is not from a published
>>> source.
>>> >         > > > Otherwise, it is adding denormalization. If this is the
>>> >         > > case, wouldn't
>>> >         > > > another name be more appropriate than "Author" (e.g.,
>>> >         > > "Quoted person")
>>> >         > > > to distinguish it from "Author" of quoted publications?
>>> >         > > >
>>> >         > > > -Ed _______________________________________________
>>> >         > > > Data-modeling mailing list
>>> >         > > > Data-modeling at freebase.com
>>> >         <mailto:Data-modeling at freebase.com>
>>> >         > > > http://lists.freebase.com/mailman/listinfo/data-modeling
>>> >         > >
>>> >         > > --
>>> >         > > Kirrily Robert
>>> >         > > Freebase Community Director
>>> >         > > kirrily at metaweb.com <mailto:kirrily at metaweb.com>
>>> >         > > http://freebase.com/
>>> >         > >
>>> >         > >
>>> >         > >
>>> >         > >
>>> >         > > _______________________________________________
>>> >         > > Data-modeling mailing list
>>> >         > > Data-modeling at freebase.com <mailto:
>>> Data-modeling at freebase.com>
>>> >         > > http://lists.freebase.com/mailman/listinfo/data-modeling
>>> >         > >
>>> >         >
>>> >         > _______________________________________________
>>> >         > Data-modeling mailing list
>>> >         > Data-modeling at freebase.com <mailto:
>>> Data-modeling at freebase.com>
>>> >         > http://lists.freebase.com/mailman/listinfo/data-modeling
>>> >         >
>>> >         > _______________________________________________
>>> >         > Data-modeling mailing list
>>> >         > Data-modeling at freebase.com <mailto:
>>> Data-modeling at freebase.com>
>>> >         > http://lists.freebase.com/mailman/listinfo/data-modeling
>>> >
>>> >         _______________________________________________
>>> >         Data-modeling mailing list
>>> >         Data-modeling at freebase.com <mailto:Data-modeling at freebase.com>
>>>  >         http://lists.freebase.com/mailman/listinfo/data-modeling
>>> >
>>> >
>>> >
>>> >
>>> ------------------------------------------------------------------------
>>> >
>>> > _______________________________________________
>>> > Data-modeling mailing list
>>> > Data-modeling at freebase.com
>>> > http://lists.freebase.com/mailman/listinfo/data-modeling
>>>
>>> _______________________________________________
>>> Data-modeling mailing list
>>> Data-modeling at freebase.com
>>> http://lists.freebase.com/mailman/listinfo/data-modeling
>>>
>>
>>
>> _______________________________________________
>> Data-modeling mailing list
>> Data-modeling at freebase.com
>> http://lists.freebase.com/mailman/listinfo/data-modeling
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freebase.com/pipermail/data-modeling/attachments/20081203/3911e2d0/attachment-0001.htm 


More information about the Data-modeling mailing list