Thanks to Fred for getting things moving. I'll post some background information now, since I'll away all next week, and because I've been working on this problem for awhile.
First, about me: I am a publishing scholar who works in human geography, at the border of the social sciences, humanties, and with rather eclectic interests in cultural politics and media, law, and politics.
This is important to keep in mind, because I have quite practical reasons for having a history of being really frustrated with bibliographic tools and data models. You can read my blog [1] archives to get more on this, but the short version is that most existing formats for citation data are produced by and for people from the hard sciences, who typically only cite secondary data: primarily articles, conference papers, and so forth.
In the humanities, law and social sciences, however, needs are much more demanding, in part because we often cite primary data (speeches, archival documents, press releases, etc) or funky things like republished and translated books, and so forth. Put simply, I often find myself not being able to represent my data in applications.
The work I have been doing on the biblio ontology [2] is my attempt to solve these frustrations. I think it is pretty far along, even if while I consider myself an expert in bibliographic data needs for scholars (with a good understanding of library perspectives as well), I am not exactly an expert in RDF. Still I have had help from people who are (Ian Davis, Leigh Dodds, Dan Brickley, etc.).
Finally, I am also co-project lead for the OpenOffice bibliographic project, and a primary member of the OpenDocument metadata subcommittee, where I have been working for the past year or so (with people like Elias Torres) on bringing RDF support to ODF so that citation support can make use of it.
I have also put a lot of work into figuring out how to format citations for publication, which is very, very important for scholars. In fact, I formatted my recent book using this sytem.
The goals reflected my work on the ontology thusfar are:
1) should be a superset of legacy formats like BibTeX, RIS, and so forth
2) must support the most demanding needs in the social sciences, humanities, and law (though the last may get a bit beyond scope, since it's another world of sorts), and those who deal with non-Western languages
3) the class system must be able to map to the type system in the citation style language I designed [3]. In short, it is not enough to just encode the data: it needs to be able to be formatted according to the often archaic details of citation styles
4) should be developer-friendly; I consider examples like DOAP and SKOS to be models here
Behind all of these goals are a more concrete goal: it should be perfect for using in OpenDocument/OpenOffice citation support.
Non-goals
1) suporting (in the sense of say, lossless mapping to) library conventions as represented in MARC
Primary intended users are scholars, researchers, students, and so forth. However, certainly it can and should be easily useable for the sort of data provide by book sellers, libraries and so forth.
The ontology in my SVN [4] is much more comprehensive than what is published. Primary issues I was seeing are:
i. the classes (need to settle them and their mapping to the CSL style conventions) ii.contributor modeling (a complicated problem because of the signicance of order in bibliographic references, though I think I know how to solve it)
Recent experiements are looking better to me [5].
Sorry for the brain-dump, but it ought to give people plenty to chew on while I'm gone ;-)
> Thanks to Fred for getting things moving. I'll post some background > information now, since I'll away all next week, and because I've been > working on this problem for awhile.
Perfect, this is exactly what I was looking for.
> First, about me: I am a publishing scholar who works in human > geography, at the border of the social sciences, humanties, and with > rather eclectic interests in cultural politics and media, law, and > politics.
> This is important to keep in mind, because I have quite practical > reasons for having a history of being really frustrated with > bibliographic tools and data models. You can read my blog [1] archives > to get more on this, but the short version is that most existing > formats for citation data are produced by and for people from the hard > sciences, who typically only cite secondary data: primarily articles, > conference papers, and so forth.
Good, so there are motivations :)
> In the humanities, law and social sciences, however, needs are much > more demanding, in part because we often cite primary data (speeches, > archival documents, press releases, etc) or funky things like > republished and translated books, and so forth. Put simply, I often > find myself not being able to represent my data in applications.
This is why I am proposing the same approach as we did with the music ontology: multi-expressiveness-levels.
Anyway, if you have the time, take a look at the music ontology, there are many similarities between the two ontologies (not the same domain, but many similar concepts, same problems and issues, same sort of goals and users, etc.)
> The work I have been doing on the biblio ontology [2] is my attempt to > solve these frustrations. I think it is pretty far along, even if > while I consider myself an expert in bibliographic data needs for > scholars (with a good understanding of library perspectives as well), > I am not exactly an expert in RDF. Still I have had help from people > who are (Ian Davis, Leigh Dodds, Dan Brickley, etc.).
This is the power of a group or a community: using the strength of everyone!
> Finally, I am also co-project lead for the OpenOffice bibliographic > project, and a primary member of the OpenDocument metadata > subcommittee, where I have been working for the past year or so (with > people like Elias Torres) on bringing RDF support to ODF so that > citation support can make use of it.
Fantastic, it will be a great test case for the ontology.
> 2) must support the most demanding needs in the social sciences, > humanities, and law (though the last may get a bit beyond scope, since > it's another world of sorts), and those who deal with non-Western > languages
Well, this is why a modular approach should be use. That way, if the ontology is well done, we will be able to easily plug extension modules to describe more precision and complexe things like the law things. We shouldn't burden the "core" ontology with such precise things, however we should enable the possibility to use such an extension (another time, same minding as the music ontology.
> 3) the class system must be able to map to the type system in the > citation style language I designed [3]. In short, it is not enough to > just encode the data: it needs to be able to be formatted according to > the often archaic details of citation styles
> 4) should be developer-friendly; I consider examples like DOAP and > SKOS to be models here
definitely > Behind all of these goals are a more concrete goal: it should be > perfect for using in OpenDocument/OpenOffice citation support.
> Non-goals
> 1) suporting (in the sense of say, lossless mapping to) library > conventions as represented in MARC
> Primary intended users are scholars, researchers, students, and so > forth. However, certainly it can and should be easily useable for the > sort of data provide by book sellers, libraries and so forth.
Exact.
Great, thanks for taking the time to write that great introduction. From there, we will be able to build-up and ultimately to create that ontology :0
On Apr 15, 11:30 am, Frederick Giasson <f...@fgiasson.com> wrote:
...
> 2) must support the most demanding needs in the social sciences, > > humanities, and law (though the last may get a bit beyond scope, since > > it's another world of sorts), and those who deal with non-Western > > languages
> Well, this is why a modular approach should be use. That way, if the > ontology is well done, we will be able to easily plug extension modules > to describe more precision and complexe things like the law things. We > shouldn't burden the "core" ontology with such precise things, however > we should enable the possibility to use such an extension (another time, > same minding as the music ontology.
I definitely believe it is possible and important to create something pretty clean and simple but also flexible. The key is really the relationship properties.
FWIW, in the SVN version of the ontology, I was experimenting with an idea that I think has a lot of promise. I hooked properties like author, editor and so forth into the FRBR ontology (author subproperty of frbr:creator, editor of frbr:realizer and so forth) with the idea that more complex relationships could be inferred without forcing people to have to explicitly model FRBR.
Just an idea ...
BTW, I did look at the music ontology earlier, in part because of the FRBR connection. I actually like FRBR, even if I don't want to actually have to encode all the complicated relationships (it gets messy for a lot of citation stuff)!
