Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-owl] Best of Breed OWL API

Cedric Peeters cedric.peeters at fundp.ac.be
Wed Mar 25 06:57:05 PDT 2009


Hello,

Maybe you can have a look at this paper where there is a comparison of  
some APIs : http://lsdis.cs.uga.edu/library/download/JK05-ISWC2005.pdf
In this paper you can find references to datasets for benchmarking.

You can also take a look at 2 other APIs : WOLAPI and kaon2.

For my work, I've chosen OWLAPI because after a few tests (also with  
Jena and Kaon2) it was the fastest API for my ontology and work.

Hope it helps,

Cedric



Nick Khamis <symack at gmail.com> a écrit :

> Forwarded conversation
> Subject: Best of Breed OWL API
> ------------------------
>
> From: *Nick Khamis* <symack at gmail.com>
> Date: Fri, Mar 20, 2009 at 11:13 PM
> To: topbraid-composer-users at googlegroups.com, GATE ML <
> gate-users at lists.sourceforge.net>
>
>
> Hello Everyone,
>
> We are currently considering one of the following (jena.jar, gate.jar,
> protege.jar) libraries to assist in CRUD (Create, Read, Update, Delete)
> operations of ontologies. We will be using triple stores such as (jenadb,
> sdb, rdfcrm). I am trying to learn from your experiences regarding this
> issue and what  your advice (pros and cons) is in terms of using the listed
> libraries. We are also open to recommendations of other APIs that I have not
> listed.
>
> Regards,
> Ninus.
>
> ----------
> From: *Jeremy Carroll* <jeremy at topquadrant.com>
> Date: Fri, Mar 20, 2009 at 11:35 PM
> To: topbraid-composer-users at googlegroups.com, GATE ML <
> gate-users at lists.sourceforge.net>
>
>
>
>
> Please clarify: CRUD operations for
>
> -          ontologies
>
> -          triples in ontologies expressed in RDF
>
> -          or something else? (e.g. axioms in ontologies, or individuals in
> say an OWL Ontology)
>
>
>
>
>
> Concerning Jena, when I worked on that project, the belief was that the
> typical life cycle had
>
>    Create + Update operations, followed by Read operations
>
> and we tended to design with that in mind, while supporting CRUD in any
> order. i.e. we believed Jena was being used more for analysis type
> applications than OLTP type applications.
>
>
>
> Of course, I don?t know any better than anyone else (not in the Jena team)
> what the current design goals are.
>
>
>
> Jeremy
>
>
>
>
>
>
>
> *From:* topbraid-composer-users at googlegroups.com [mailto:
> topbraid-composer-users at googlegroups.com] *On Behalf Of *Nick Khamis
> *Sent:* Friday, March 20, 2009 8:14 PM
> *To:* topbraid-composer-users at googlegroups.com; GATE ML
> *Subject:* [tbc-users] Best of Breed OWL API
>
> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups
> "TopBraid Composer Users" group.
> To post to this group, send email to
> topbraid-composer-users at googlegroups.com
> To unsubscribe from this group, send email to
> topbraid-composer-users+unsubscribe at googlegroups.com<topbraid-composer-users%2Bunsubscribe at googlegroups.com>
> For more options, visit this group at
> http://groups.google.com/group/topbraid-composer-users?hl=en
> -~----------~----~----~----~------~----~------~--~---
>
>
> ----------
> From: *Atanas Kiryakov* <naso at sirma.bg>
> Date: Sat, Mar 21, 2009 at 7:10 AM
> To: topbraid-composer-users at googlegroups.com, GATE ML <
> gate-users at lists.sourceforge.net>, Hamish Cunningham <hamish at dcs.shef.ac.uk>,
> Arjohn Kampman <arjohn.kampman at aduna-software.com>
>
>
>  Dear Nick,
>
> what you refer as "GATE OWL API" is probably the API of Sesame 1.2. Actually
> GATE integrates SwiftOWLIM 2.9.1 as part of its standard distribution (the
> same holds for TBC, as far as I know). OWLIM in general provides storage and
> reasoning backend through the Sesame APIs
>
> Speaking of APIs, Sesame's main APIs allow manipulation of RDF store. One
> can CRD RDF triples (updating RDF triple is a non-sense). Thus you can
> assert:
>
>     C1 rdfs:subClassOf C2
>
> or delete this statement or ask for any statements paterns (e.g. C1
> rdfs:subClassOf ?x). And evaluate queries, of course. This is obviously a
> low-level RDF API, that is designed for maximum efficiency when it comes to
> moving around large volumes of data in RDF (plenty of public benchmark
> results prove it is the best for this).
>
> If you want to use OO-style API, where you deal with classes and objects
> (instead of triples), you can look at the Graph API of Sesame (
> www.openrdf.org) and Elmo. There is also a simple OO-style API as part of
> GATE
>
> Finally, there are APIs which allow one to deal with OWL definitions and
> primitives (e.g. having specific programmatic support for
> owl:minCardinality), e.g. http://owlapi.sourceforge.net/. It's my personal
> opinoin that most of the users and applications (other than ontology
> enditors) usually do not need such APIs - people edit ontologies in editors
> as TBC, then only load them programatically (as RDF files) into repositories
> such as Sesame. Of course, there are plenty of different ways to get the
> same job done, depending on your application and environment.
>
> Regards,
> Naso
>
> ----------------------------------------------------------
> Atanas Kiryakov
> CEO of Ontotext AD, http://www.ontotext.com
> Sirma Group Corp, http://www.sirma.bg
> Phone: (+359 2) 8091 555; Fax: 8090 404
> ----------------------------------------------------------
>
>
> ----------
> From: *Nick Khamis* <symack at gmail.com>
> Date: Sat, Mar 21, 2009 at 8:34 AM
> To: topbraid-composer-users at googlegroups.com
>
>
> Hey Jeremy,
>
> Thanks you so much for you response... We are going to be working with OWL
> Ontologies with axioms, individuals, and reasoning using OWL-DL Tableau
> Reasoners. I have been using TBC for over a year now and I love it... I
> never looked into whether or not TBC has an API that we can use for the said
> operations? If so how complete is it.
>
> Regards,
> Ninus
>
> ----------
> From: *Nick Khamis* <symack at gmail.com>
> Date: Sat, Mar 21, 2009 at 9:11 AM
> To: topbraid-composer-users at googlegroups.com
>
>
> Hey Atanas,
>
> Thank you so much for your response. Basically we are looking to implement
> an application that automatedly generates OWL-DL Ontologies that have
> axioms, individuals and also make use a tableau reasoner, as well as store
> them in a repository. That being said I understand that SwiftOWLIM is one of
> the fastest reasoners out there, however; as you know it is not a full
> OWL-DL tableau reasoner and cannot deduce things such as disjointness. That
> being said we are looking for an OWL API (OO Style would be nice) that is
> complete in the sense that it will allow for the CRD of "complete" OWL-DL
> KBs. I think "The OWL API at http://owlapi.sourceforge.net/" is just what
> the doctor ordered, has anyone here used the API is it any good (easy to
> use, somewhat complete, well documented, supported)?
>
> Best Regards,
> Ninus.
>
> ----------
> From: *Nick Khamis* <symack at gmail.com>
> Date: Sat, Mar 21, 2009 at 9:17 AM
> To: topbraid-composer-users at googlegroups.com
>
>
> Hey Jeremy,
>
> We are also looking to generate in an automated manner the complete DL
> Ontology (Concepts, Individuals, Properties etc..), and also make use of a
> tableau reasoner.
>
> ----------
> From: *Atanas Kiryakov* <naso at sirma.bg>
> Date: Sat, Mar 21, 2009 at 12:54 PM
> To: topbraid-composer-users at googlegroups.com
>
>
>  Hi Nick
>
> if you *really* need DL reasoning SwiftOWLIM and Sesame's APIs are not for
> you. I am emphasizing *really* because there are plenty of people who chose
> DL being influenced by something and afterwards get surprised by the
> specific constraints that it requires
>
> As a general comment, the completeness of the reasoning is not determined by
> the API (i.e. the interface definitions), but rather from the implementation
> that you will choose. For instance, one can provide results of a sound and
> complete inference through Sesame's interfaces (OWLIM could have been doing
> this, if it was a tableau reasoner) and another one could provide shallow or
> incomplete inference through DIG (yo u have to look at this, if you are into
> DL reasoning)
>
>
> ----------
> From: *Peter Crowther* <Peter.Crowther at melandra.com>
> Date: Mon, Mar 23, 2009 at 6:34 AM
> To: GATE ML <gate-users at lists.sourceforge.net>
>
>
>> From: Atanas Kiryakov [naso at sirma.bg]
> Others of us have different opinions, and regard RDF-level APIs as
> shockingly low-level and OWL APIs as "about right" :-).  But then, I'm from
> the description logic side of the world and would have preferred OWL to be a
> set of XML extensions to RDF rather than an interpretation of particular RDF
> triples.
>
> As Antanas says, there are plenty of different ways to get the same job
> done!
>
> - Peter
>
> ------------------------------------------------------------------------------
> Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
> powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
> easily build your RIAs with Flex Builder, the Eclipse(TM)based development
> software that enables intelligent coding and step-through debugging.
> Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
> _______________________________________________
> GATE-users mailing list
> GATE-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/gate-users
>
> ----------
> From: *Scott Henninger* <shenninger at topquadrant.com>
> Date: Tue, Mar 24, 2009 at 2:03 PM
> To: TopBraid Composer Users <topbraid-composer-users at googlegroups.com>
>
>
>
> Nick;  In terms of access to TBC functionality, TopBraid supports a
> couple of way to get to Composer/Live functionality.  Probably the
> most relevant in your case is SPARQLMotion (http://www.topquadrant.com/
> sparqlmotion/ <http://www.topquadrant.com/%0Asparqlmotion/>).  This has
> modules for reasoners, SPARQL queries,
> import/export formats, etc.  The scripts can be exposed as REST Web
> Services.
>
> The implication is that SPARQLMotion has access to a large set of
> Composer features and you can access this from the outside world by
> creating a script, exposing the script as a Web service, and storing
> it on TopBraid Live.  This is quite powerful, as you can design a
> series of ontology manipulations using a graphical language and access
> this via a simple Web service call.  In many ways, the API is yours to
> design.
>
> Ensemble also provides access to Composer features via a configurable
> Web application.
>
> -- Scott
>> On Fri, Mar 20, 2009 at 11:35 PM, Jeremy Carroll <jer... at topquadrant.com
>> wrote:
>
> ----------
> From: *Nick Khamis* <symack at gmail.com>
> Date: Wed, Mar 25, 2009 at 8:36 AM
> To: topbraid-composer-users at googlegroups.com
>
>
> Hello Everyone,
>
> So just to wrap up this thread, and this information may be useful for
> others...
>
> When talking about creating an OWL-DL ontology editing application using
> java code and an Ontology API to help do so we have:
>
> Name
> Link                                                           Desc
> The OWL API                        http://owlapi.sourceforge.net/.
> OO Style OWL-DL API
> Jena                                     http://jena.sourceforge.net/
>
> Jeremy you brought up the comment that Jena was designed more for analysis
> type applications as opposed to ontology editing applications. From the
> little that I have experimented (jena java code examples) I saw that Jena
> could be a good candidate for use as an API for an ontology editing
> application? I put a question mark because I would like your expert
> elaboration on this please.
>
> We also have a different option as brought up by Scott. Using TBC's API to
> edit ontologies using SPARQLMotion and web services (interesting take).
>
> We talked a little about how to integrate the reasoning services, into an
> ontology editing application... Atanas mentioned DIG and also pointed out
> how we could have "shallow" inferencing when using DIG. I did experiment and
> read up on DIG and also agree that is not a good idea, furthermore; the
> ontology editing application will reside on the same server as the reasoner
> making less need for using DIG.... That being said what are your takes on
> how to integrate reasoning services using java code and (new topic :) a
> "Reasoning Services API".
> If I remember correctly TBC uses DIG for pellet and other reasoners that we
> may want to add, the new version of Protege stopped using DIG for reasoning
> services the last time I looked.
>
> I would like to thank everyone in advanced for your expertise and input,
>
> Regards,
> Ninus
>






More information about the protege-owl mailing list