A query about RDF statements

48 views
Skip to first unread message

Nick Bywell

unread,
May 16, 2013, 8:31:50 AM5/16/13
to island...@googlegroups.com
Hi,

Can anyone advise on the appropriate course of action when there are no existing predicates in any RDF specifications that meet one's needs please?

Thanks,

Nick Bywell

The Freshwater Biological Association

Nigel Banks

unread,
May 16, 2013, 9:28:05 AM5/16/13
to island...@googlegroups.com
I worked on a document about this very subject not long ago with the help of Donald Moses, I stick the relevant chunk inline here:

Fedora Ontology

http://www.fedora.info/definitions/1/0/fedora-relsext-ontology.rdfs


We have the fedora ontology for representing relationships between objects. The ontology is hierarchical where all relationships derive from the generic fedora:fedoraRelationship.


  • fedora:fedoraRelationship

    • isPartOf

      • isConstituentOf

      • isMemberOf

        • isSubsetOf

        • isMemberOfCollection

    • hasPart

      • hasConstituent

      • hasMember

        • hasSubset

        • hasCollectionMember

    • isDerivationOf

    • hasDerivation

    • isDependentOf

    • hasDependent

    • HasDescription

      • isMetadataFor

      • HasMetadata

      • isAnnotationOf

      • HasAnnotation

    • hasEquivalent

Use Existing Ontologies


In general when creating a new relationship within the Islandora framework we should investigate whether or not existing ontologies can be used to represent the relationship desired. Two ontology examples from the bibliographic world include FABIO and BIBO.  

FABIO

http://vocab.ox.ac.uk/fabio


FaBiO, the FRBR-aligned Bibliographic Ontology, is an ontology for recording and publishing on the Semantic Web descriptions of entities that are published or potentially publishable, and that contain or are referred to by bibliographic references, or entities used to define such bibliographic references. FaBiO entities are primarily textual publications such as books, magazines, newspapers and journals, and items of their content such as poems and journal articles. However, they also include datasets, computer algorithms, experimental protocols, formal specifications and vocabularies, legal records, governmental papers, technical and commercial reports and similar publications, and also bibliographies, reference lists, library catalogues and similar collections. FaBiO classes are structured according to the FRBR schema of Works, Expressions, Manifestations and Items. Additional properties have been added to extends the FRBR data model by linking Works and Manifestations (fabio:hasManifestation and fabio:isManifestationOf), Works and Items (fabio:hasPortrayal and fabio:isPortrayedBy), and Expressions and Items (fabio:hasRepresentation and fabio:isRepresentedBy).

BIBO

http://bibliontology.com/specification


The Bibliographic Ontology Specification provides main concepts and properties for describing citations and bibliographic references (i.e. quotes, books, articles, etc) on the Semantic Web.


Depending on the domain, there is probably an ontology that includes the relationship sought.  It would be a best practice to integrate with an existing ontology rather than create our own ‘siloed’ relationships within Islandora.

Should we define custom non literal relationships?


This section is meant to investigate whether or not there are any tangible benefits for developers or repository managers in refining the relationships between objects; by adding more specific relationships such as islandora:isIssueOf or islandora:isPageOf between objects in addition to existing relationships such as fedora:isMemberOf.


I originally thought that expressing specific relationships along with the generic fedora:isMemberOf, would allow us to build more complicated interfaces while still having a basic set of relationships for structural navigation. On closer examination, we are not limited by the set of fedora structural relationships. As we can limit our query’s to specific content models to get the same effect.


For example by using islandora:isIssueOf we express the relationship between two objects and also the type of the object in the triple. We can express this equivalently by using fedora:isMemberOf and by limiting our query to objects that have a content model of type islandora:newspaperIssueCModel. It is a slightly longer query but nonetheless it’s more flexible and require less maintenance for developers and repository managers.


Arguably its a bad idea to construct new relationship statements based on the type of the object in the triple. It results in more RDF statements that need to be maintained and it does not add any information to the repository that was not already present.


tldr; No

Nicholas Bywell

unread,
May 17, 2013, 5:18:29 AM5/17/13
to island...@googlegroups.com

Thanks Nigel. This has helped focus my thoughts on this.

 

Nick

 

From: island...@googlegroups.com [mailto:island...@googlegroups.com] On Behalf Of Nigel Banks
Sent: 16 May 2013 14:28
To: island...@googlegroups.com
Subject: [islandora-dev] Re: A query about RDF statements

 

I worked on a document about this very subject not long ago with the help of Donald Moses, I stick the relevant chunk inline here:

Fedora Ontology

http://www.fedora.info/definitions/1/0/fedora-relsext-ontology.rdfs

 

We have the fedora ontology for representing relationships between objects. The ontology is hierarchical where all relationships derive from the generic fedora:fedoraRelationship.

 

·         fedora:fedoraRelationship

o    isPartOf

§  isConstituentOf

§  isMemberOf

·         isSubsetOf

·         isMemberOfCollection

o    hasPart

§  hasConstituent

§  hasMember

·         hasSubset

·         hasCollectionMember

o    isDerivationOf

o    hasDerivation

o    isDependentOf

o    hasDependent

o    HasDescription

§  isMetadataFor

§  HasMetadata

§  isAnnotationOf

§  HasAnnotation

o    hasEquivalent

Use Existing Ontologies

 

In general when creating a new relationship within the Islandora framework we should investigate whether or not existing ontologies can be used to represent the relationship desired. Two ontology examples from the bibliographic world include FABIO and BIBO.  

FABIO

http://vocab.ox.ac.uk/fabio

 

FaBiO, the FRBR-aligned Bibliographic Ontology, is an ontology for recording and publishing on the Semantic Web descriptions of entities that are published or potentially publishable, and that contain or are referred to by bibliographic references, or entities used to define such bibliographic references. FaBiO entities are primarily textual publications such as books, magazines, newspapers and journals, and items of their content such as poems and journal articles. However, they also include datasets, computer algorithms, experimental protocols, formal specifications and vocabularies, legal records, governmental papers, technical and commercial reports and similar publications, and also bibliographies, reference lists, library catalogues and similar collections. FaBiO classes are structured according to the FRBR schema of Works, Expressions, Manifestations and Items. Additional properties have been added to extends the FRBR data model by linking Works and Manifestations (fabio:hasManifestation and fabio:isManifestationOf), Works and Items (fabio:hasPortrayal and fabio:isPortrayedBy), and Expressions and Items (fabio:hasRepresentation and fabio:isRepresentedBy).

BIBO

http://bibliontology.com/specification

 

The Bibliographic Ontology Specification provides main concepts and properties for describing citations and bibliographic references (i.e. quotes, books, articles, etc) on the Semantic Web.

 

Depending on the domain, there is probably an ontology that includes the relationship sought.  It would be a best practice to integrate with an existing ontology rather than create our own ‘siloed’ relationships within Islandora.

Should we define custom non literal relationships?

 

This section is meant to investigate whether or not there are any tangible benefits for developers or repository managers in refining the relationships between objects; by adding more specific relationships such as islandora:isIssueOf or islandora:isPageOf between objects in addition to existing relationships such as fedora:isMemberOf.

 

I originally thought that expressing specific relationships along with the generic fedora:isMemberOf, would allow us to build more complicated interfaces while still having a basic set of relationships for structural navigation. On closer examination, we are not limited by the set of fedora structural relationships. As we can limit our query’s to specific content models to get the same effect.

 

For example by using islandora:isIssueOf we express the relationship between two objects and also the type of the object in the triple. We can express this equivalently by using fedora:isMemberOf and by limiting our query to objects that have a content model of type islandora:newspaperIssueCModel. It is a slightly longer query but nonetheless it’s more flexible and require less maintenance for developers and repository managers.

 

Arguably its a bad idea to construct new relationship statements based on the type of the object in the triple. It results in more RDF statements that need to be maintained and it does not add any information to the repository that was not already present.


tldr; No

On Thursday, May 16, 2013 2:31:50 PM UTC+2, Nick Bywell wrote:

Hi,

Can anyone advise on the appropriate course of action when there are no existing predicates in any RDF specifications that meet one's needs please?

Thanks,

Nick Bywell

The Freshwater Biological Association

--
You received this message because you are subscribed to the Google Groups "islandora-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to islandora-de...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 



The Freshwater Biological Association

The Ferry Landing, Far Sawrey, Ambleside, Cumbria, LA22 0LP, United Kingdom
Tel. +44 (0)1539 442468 Fax. +44 (0)1539 446914 Email. in...@fba.org.uk
Website. www.fba.org.uk Director: Dr Michael Dobson
Registered Charity No. 214440 Company Limited by Guarantee No. 263162, England UKPRN No. 10018314, VAT No. GB 153895238


If you have received this message by mistake, please notify the sender immediately, delete it and do not copy it to anyone else.
This communication may contain privileged or confidential information which is exempt from disclosure under English law and this information may not be used or disclosed except for the purpose for which it has been sent. Any views expressed by the sender of this email are not necessarily those of the FBA. FBA will not accept any liability in respect of such a communication, and the employee responsible will be personally liable for any damages or other liabilities arising.
Computer viruses can be transmitted by email and you should be aware that emails may be intercepted by third parties. You are advised to check this email and any attachments for the presence of viruses as neither FBA nor the sender accept responsibility for any viruses transmitted by this email and/or any attachments.
The FBA does not generally engage in systematic monitoring activities although it reserves the right to do so where there is reason to believe that misuse of its computing facilities is occurring.

Don

unread,
May 19, 2013, 10:57:59 AM5/19/13
to island...@googlegroups.com
Nick:
is there a particular type of relationship you want to create ? The examples Nigel points to are primarily for bibliographic data.  If you had a science-based need there may be other ontologies that would contain the relationship you were looking for.
Don

Giancarlo Birello

unread,
May 19, 2013, 11:17:31 AM5/19/13
to island...@googlegroups.com
Nick, I was playing around ontology, starting from a module of Rosie, you can found it here https://github.com/giancarlobi/islandora_relationship_editor
I used AOS AVOC ontology from FAO, as suggested by Mark, but there are a lot of OWL ready ontology for any kind of science (i.e. Tones repository http://owl.cs.manchester.ac.uk/repository/)
Best,
Giancarlo

Nicholas Bywell

unread,
May 21, 2013, 5:50:45 AM5/21/13
to island...@googlegroups.com

Hi Don,

 

Examples, relating to the relationship between taxon objects would be “hasBasionym”, “hasAcceptedName”, “hasSynonym”. We could perhaps use MADS-RDF, but we are already using such terms available in that schema for our vocabulary. This obviously does not preclude their re use for this purpose, but it would not be ideal.

 

In the case where we have taxonomic information that is not accommodated by Darwin Core, such as specific info about identification codes for chrysophytes and their dimensions, our temporary solution is to store this specific information in a separate object with a pseudo-Darwin Core datastream. The RDF statements linking objects that contain the valid-Darwin Core objects (representing a particular taxon) with the objects that contain the pseudo-Darwin Core datastream would perhaps have predicates of “isExtendedBy” and “extends”.

 

Any advice relating to the RDF issue or the broader issues raised by all this would be gratefully received.

 

Nick

 

From: island...@googlegroups.com [mailto:island...@googlegroups.com] On Behalf Of Don
Sent: 19 May 2013 15:58
To: island...@googlegroups.com
Subject: [islandora-dev] Re: A query about RDF statements

 

Nick:

--

You received this message because you are subscribed to the Google Groups "islandora-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to islandora-de...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Nicholas Bywell

unread,
May 21, 2013, 6:02:18 AM5/21/13
to island...@googlegroups.com

Hi Don/Giancarlo,

 

Sorry Giancarlo, I had not seen your response when I replied to Don’s post. I am re-posting to retain all the info within the one thread. I will look into the links you kindly supplied.

 

Examples, relating to the relationship between taxon objects would be “hasBasionym”, “hasAcceptedName”, “hasSynonym”. We could perhaps use MADS-RDF, but we are already using such terms available in that schema for our vocabulary. This obviously does not preclude their re use for this purpose, but it would not be ideal.

 

In the case where we have taxonomic information that is not accommodated by Darwin Core, such as specific info about identification codes for chrysophytes and their dimensions, our temporary solution is to store this specific information in a separate object with a pseudo-Darwin Core datastream. The RDF statements linking objects that contain the valid-Darwin Core objects (representing a particular taxon) with the objects that contain the pseudo-Darwin Core datastream would perhaps have predicates of “isExtendedBy” and “extends”.

 

Any advice relating to the RDF issue or the broader issues raised by all this would be gratefully received.

 

Nick

 

 

From: island...@googlegroups.com [mailto:island...@googlegroups.com] On Behalf Of Giancarlo Birello
Sent: 19 May 2013 16:18
To: island...@googlegroups.com
Subject: [islandora-dev] Re: A query about RDF statements

 

Nick, I was playing around ontology, starting from a module of Rosie, you can found it here https://github.com/giancarlobi/islandora_relationship_editor

--

You received this message because you are subscribed to the Google Groups "islandora-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to islandora-de...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Reply all
Reply to author
Forward
0 new messages