Modifications and things to consider.

5 views
Skip to first unread message

Frederick Giasson

unread,
Jul 25, 2007, 10:20:16 AM7/25/07
to bibliographic-ontolog...@googlegroups.com
Hi all,


there is a list of things I fixed, and things to consider.

- Added dcterms:publisher in the external ontologies properties we reuse

- Added: bibo:producer, bibo:distributor and bibo:owner. Remember that
we deleted these roles because we want to make a distinction between the
a Work (frbr) and its Manifestation or Itemisation (the publishing of a
work, or its owning).

- Removed dcterms:type since we moved away of the "types" (individual)
methods for typing documents.

- Changed the rdfs:domain of bibo:uri to rdfs:Resource (instead of
rdfs:Literal). Think it make sense. So, bibo:uri will be able to
identify the resource that identifies the document... humm... should be
discussed, this seem redundant :) ) Also, I restricted the domain do
bibo:Document and bibo:Collection. However, should we open the domain of
this property to rdfs:Resource? So that anyone can reuse it?

- Added bibo:adr and bibo:locality (Bruce, could you write the
description for both of them please?)

- Fixed examples on the wiki related to bibo:url and how we handle web pages


Yet to be discussed before inclusion:

- bibo:organisation and/or bibo:department and/or bibo:sponsor. what
would be the definition and use of bibo:organisation and
bibo:department. Also, I would use one of these two and bibo:sponsor.
Bibo:sponsor would be use for events (could be a sub-property of
even:hasAgent). But would it makes sense to use bibo:sponsor for
Documents as well? Think so since documents (or collection?) writings
could be sponsored, no?

- bibo:sentFrom and bibo:sentTo and all the related stuff (related to
contributions and letters publishing).

- Should we keep bibo:url subproperty of bibo:uri? I am still not sure
how to model that. bibo:uri seems redundant if a Document or and Article
is a Web document.

- Would need another example of web page and internet site. We should
find a law document on the web, and then describing it with bibo (as
well as its collection, so the internet site).

- Should put some more example of how DCTERMS can be reused
(dcterms:format, licence, etc).


Take care,


Fred

Bruce D'Arcus

unread,
Jul 25, 2007, 2:24:05 PM7/25/07
to bibliographic-ontolog...@googlegroups.com
Frederick Giasson wrote:

> - Added: bibo:producer, bibo:distributor and bibo:owner. Remember that
> we deleted these roles because we want to make a distinction between the
> a Work (frbr) and its Manifestation or Itemisation (the publishing of a
> work, or its owning).

Should we link these to the FRBR analogs?

> - Removed dcterms:type since we moved away of the "types" (individual)
> methods for typing documents.

We might still need this or something similar in some cases though.

> - Changed the rdfs:domain of bibo:uri to rdfs:Resource (instead of
> rdfs:Literal). Think it make sense. So, bibo:uri will be able to
> identify the resource that identifies the document... humm... should be
> discussed, this seem redundant :) ) Also, I restricted the domain do
> bibo:Document and bibo:Collection. However, should we open the domain of
> this property to rdfs:Resource? So that anyone can reuse it?
>
> - Added bibo:adr and bibo:locality (Bruce, could you write the
> description for both of them please?)

I think we can use vcard now for this, as well as names.

...

> Yet to be discussed before inclusion:
>
> - bibo:organisation and/or bibo:department and/or bibo:sponsor. what
> would be the definition and use of bibo:organisation and
> bibo:department. Also, I would use one of these two and bibo:sponsor.
> Bibo:sponsor would be use for events (could be a sub-property of
> even:hasAgent). But would it makes sense to use bibo:sponsor for
> Documents as well? Think so since documents (or collection?) writings
> could be sponsored, no?
>
> - bibo:sentFrom and bibo:sentTo and all the related stuff (related to
> contributions and letters publishing).

I'm still thinking about all this.

> - Should we keep bibo:url subproperty of bibo:uri? I am still not sure
> how to model that. bibo:uri seems redundant if a Document or and Article
> is a Web document.
>
> - Would need another example of web page and internet site. We should
> find a law document on the web, and then describing it with bibo (as
> well as its collection, so the internet site).
>
> - Should put some more example of how DCTERMS can be reused
> (dcterms:format, licence, etc).

I'll see if I can find some time, but am busy with a lot of other
(mostly stupid!) things.

Bruce

Frederick Giasson

unread,
Jul 25, 2007, 5:36:24 PM7/25/07
to bibliographic-ontolog...@googlegroups.com
Hi Bruce,

>> - Added: bibo:producer, bibo:distributor and bibo:owner. Remember that
>> we deleted these roles because we want to make a distinction between the
>> a Work (frbr) and its Manifestation or Itemisation (the publishing of a
>> work, or its owning).
>>
>
> Should we link these to the FRBR analogs?
>
>

Yes, I think so.

>> - Removed dcterms:type since we moved away of the "types" (individual)
>> methods for typing documents.
>>
>
> We might still need this or something similar in some cases though.
>
>

Well, people can always re-use it if they want, but if we don't have any
usecase for now with the ontology, I would get it out of the
specification's document for now.

>> - Changed the rdfs:domain of bibo:uri to rdfs:Resource (instead of
>> rdfs:Literal). Think it make sense. So, bibo:uri will be able to
>> identify the resource that identifies the document... humm... should be
>> discussed, this seem redundant :) ) Also, I restricted the domain do
>> bibo:Document and bibo:Collection. However, should we open the domain of
>> this property to rdfs:Resource? So that anyone can reuse it?
>>
>> - Added bibo:adr and bibo:locality (Bruce, could you write the
>> description for both of them please?)
>>
>
> I think we can use vcard now for this, as well as names.
>

Well, check the geoname mail thread; now I am not so sure; I think we
should create a bibo:locality that really explains the goal of this
property, and to bind it with a SpacialThing so that we can use it with
any place descriptions.

>> Yet to be discussed before inclusion:
>>
>> - bibo:organisation and/or bibo:department and/or bibo:sponsor. what
>> would be the definition and use of bibo:organisation and
>> bibo:department. Also, I would use one of these two and bibo:sponsor.
>> Bibo:sponsor would be use for events (could be a sub-property of
>> even:hasAgent). But would it makes sense to use bibo:sponsor for
>> Documents as well? Think so since documents (or collection?) writings
>> could be sponsored, no?
>>
>> - bibo:sentFrom and bibo:sentTo and all the related stuff (related to
>> contributions and letters publishing).
>>
>
> I'm still thinking about all this.
>
>

Perfect, looking forward that.


>> - Should we keep bibo:url subproperty of bibo:uri? I am still not sure
>> how to model that. bibo:uri seems redundant if a Document or and Article
>> is a Web document.
>>
>> - Would need another example of web page and internet site. We should
>> find a law document on the web, and then describing it with bibo (as
>> well as its collection, so the internet site).
>>
>> - Should put some more example of how DCTERMS can be reused
>> (dcterms:format, licence, etc).
>>
>
> I'll see if I can find some time, but am busy with a lot of other
> (mostly stupid!) things.
>
>

nice


Take care,

Fred

Bruce D'Arcus

unread,
Jul 25, 2007, 11:04:04 PM7/25/07
to bibliographic-ontolog...@googlegroups.com
OK, some ideas ...

Frederick Giasson wrote:

> Yet to be discussed before inclusion:
>
> - bibo:organisation and/or bibo:department and/or bibo:sponsor. what
> would be the definition and use of bibo:organisation and
> bibo:department. Also, I would use one of these two and bibo:sponsor.
> Bibo:sponsor would be use for events (could be a sub-property of
> even:hasAgent). But would it makes sense to use bibo:sponsor for
> Documents as well? Think so since documents (or collection?) writings
> could be sponsored, no?

On this, for events we need something that specifies the organizations
(or maybe groups) that organize or host conferences, hearings (some
committee in the U.S. Senate say), workshops, etc.

I agree with Ivan on second thought that sponsor probably isn't quite right.

Maybe something like "organizer"?

<http://dev.nepomuk.semanticdesktop.org/repos/trunk/ontologies/nie/htmldocs/ncal.html#organizer>

Still thinking on the thesis issue :-)

> - bibo:sentFrom and bibo:sentTo and all the related stuff (related to
> contributions and letters publishing).

Why not keep this one simple and just do bibo:place? The to/from stuff
isn't necessary because it can be inferred from the role.

The same property could then also be used for publishers?

Bruce

Frederick Giasson

unread,
Jul 26, 2007, 9:10:14 AM7/26/07
to bibliographic-ontolog...@googlegroups.com
Hi Bruce,

>> Yet to be discussed before inclusion:
>>
>> - bibo:organisation and/or bibo:department and/or bibo:sponsor. what
>> would be the definition and use of bibo:organisation and
>> bibo:department. Also, I would use one of these two and bibo:sponsor.
>> Bibo:sponsor would be use for events (could be a sub-property of
>> even:hasAgent). But would it makes sense to use bibo:sponsor for
>> Documents as well? Think so since documents (or collection?) writings
>> could be sponsored, no?
>>
>
> On this, for events we need something that specifies the organizations
> (or maybe groups) that organize or host conferences, hearings (some
> committee in the U.S. Senate say), workshops, etc.
>
> I agree with Ivan on second thought that sponsor probably isn't quite right.
>
> Maybe something like "organizer"?
>
> <http://dev.nepomuk.semanticdesktop.org/repos/trunk/ontologies/nie/htmldocs/ncal.html#organizer>
>
> Still thinking on the thesis issue :-)
>
>

So, what you are suggesting is creating some properties, subPropertyOf
event:hasAgent, right? These new properties will be the specialization
of this general event property that relate the event with an Agent
(person or organization).


>> - bibo:sentFrom and bibo:sentTo and all the related stuff (related to
>> contributions and letters publishing).
>>
>
> Why not keep this one simple and just do bibo:place? The to/from stuff
> isn't necessary because it can be inferred from the role.
>
> The same property could then also be used for publishers?
>

So, bibo:place (domain bibo:Contribution?) would link a contribution
(anything else?) to a place (SpacialThing as range)?


In that case, we will add these new roles: bibo:sender and bibo:recipient ?

Salutations,


Fred

Bruce D'Arcus

unread,
Jul 26, 2007, 9:45:40 AM7/26/07
to bibliographic-ontolog...@googlegroups.com
Frederick Giasson wrote:

...

>> Maybe something like "organizer"?
>>
>> <http://dev.nepomuk.semanticdesktop.org/repos/trunk/ontologies/nie/htmldocs/ncal.html#organizer>
>>
>> Still thinking on the thesis issue :-)
>>
>>
> So, what you are suggesting is creating some properties, subPropertyOf
> event:hasAgent, right? These new properties will be the specialization
> of this general event property that relate the event with an Agent
> (person or organization).

Well, we have two existing ontologies: event and ncal. Was observing
that we can use something like ncal:organizer; not sure how best to do that.

>>> - bibo:sentFrom and bibo:sentTo and all the related stuff (related to
>>> contributions and letters publishing).
>>>
>> Why not keep this one simple and just do bibo:place? The to/from stuff
>> isn't necessary because it can be inferred from the role.
>>
>> The same property could then also be used for publishers?
>
> So, bibo:place (domain bibo:Contribution?) would link a contribution
> (anything else?)

Also an Agent.

> to a place (SpacialThing as range)?

Probably.

> In that case, we will add these new roles: bibo:sender and bibo:recipient ?

Yes, though author could be fine for sender (which is not strictly
necessary).

Bruce

Reply all
Reply to author
Forward
0 new messages