Just use this example for in-memory dataset with inference:
https://github.com/apache/jena/blob/main/jena-fuseki2/examples/config-inference-1.ttl
On 23.02.22 12:28, Luca Turchet wrote:
For the moment I don't use a TDB2, but a regular .ttl file.
Could you please send me the assembler modified to use the ttl file?
Also, is there a way to use both the --conf and --file options together?
Cheers
Luca
-----------------------------------------------------------------------------------------
*Luca Turchet*
Associate Professor
Head of the Creative, Intelligent & Multisensory Interactions Laboratory
*https://www.cimil.disi.unitn.it/* <https://www.cimil.disi.unitn.it/>
Department of Information Engineering and Computer Science
University of Trento
Via Sommarive 9 - 38123 Trento - Italy
E-mail: [email protected]
Tel: +39 0461 283792
Il giorno mer 23 feb 2022 alle ore 12:19 Luca Turchet <[email protected]>
ha scritto:
Ok received. I proceed and let you know
Luca
-----------------------------------------------------------------------------------------
*Luca Turchet*
Associate Professor
Head of the Creative, Intelligent & Multisensory Interactions Laboratory
*https://www.cimil.disi.unitn.it/* <https://www.cimil.disi.unitn.it/>
Department of Information Engineering and Computer Science
University of Trento
Via Sommarive 9 - 38123 Trento - Italy
E-mail: [email protected]
Tel: +39 0461 283792
Il giorno mer 23 feb 2022 alle ore 12:16 Lorenz Buehmann <
[email protected]> ha scritto:
To speedup the process I uploaded a tarball containing
- the extracted Fuseki 4.4.0 with the necessary Openllet Jars
- the assembler config
- please set FUSEKI_BASE then run the server with --conf assembler.ttl
- and of course modify the assembler file to link to your TDB2 location
path
Link: https://www.file.io/R2ls/download/7YKjIFR0eWyb
On 23.02.22 12:05, Lorenz Buehmann wrote:
On 23.02.22 11:54, Luca Turchet wrote:
So, firstly mvn --version provided the JDK 17.02, but the JDK version
can
be set with "export JAVA_HOME= "
I installed JDK 11 and used the POM.xml configuration you suggested.
I also
exported the JAVA_HOME in the shell session to make sure
that JDK 11 is used (with export
JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk-11.0.13.jdk/Contents/Home/).
I got only the error :
ERROR: Type 'openllet help' for usage.
what did you do here? why would you get this error? This looks more
like a commandline script call?
You should simply call (just skip the test for speedup) to build the
Openllet project:
mvn clean install -Dmaven.test.skip=true
but follow the instruction of my previous email, you have to put the
Jar file generated in openllet/distribution/target to the Fuseki
classpath
No more information in the output. I repeated with JDK 13 getting the
same
error.
it would be great to port fuseki to JDK 17 as JDK 11 is pretty old.
Luca
-----------------------------------------------------------------------------------------
*Luca Turchet*
Associate Professor
Head of the Creative, Intelligent & Multisensory Interactions
Laboratory
*https://www.cimil.disi.unitn.it/* <https://www.cimil.disi.unitn.it/>
Department of Information Engineering and Computer Science
University of Trento
Via Sommarive 9 - 38123 Trento - Italy
E-mail: [email protected]
Tel: +39 0461 283792
Il giorno mer 23 feb 2022 alle ore 10:29 Lorenz Buehmann <
[email protected]> ha scritto:
Ok, I'm still on JDK 11 and this worked for me - so you could give it
a
try. If not then can't you set the compiler plugin to 13? Although I'm
wondering why it failed with JDK 17 if you have it installed. Did you
also set this as your current JDK?
mvn --version should have shown Java 17 then
But let's try with JDK 11 first, Fuseki distribution is currently also
on Java 11
On 23.02.22 10:22, Luca Turchet wrote:
I have amended the file as you suggested (and even modifying
the aven-enforcer-plugin <version> to 3.0.0) but the result is the
same.
I attach the pom POM file.
However, I don't have installed JDK version 11, or 15. I have 13 14,
16 and 17. Do I need to install JDK version 11?
Cheers
Luca
-----------------------------------------------------------------------------------------
*Luca Turchet*
Associate Professor
Head of the Creative, Intelligent & Multisensory Interactions
Laboratory
*https://www.cimil.disi.unitn.it/* <https://www.cimil.disi.unitn.it/
Department of Information Engineering and Computer Science
University of Trento
Via Sommarive 9 - 38123 Trento - Italy
E-mail: [email protected]
Tel: +39 0461 283792
Il giorno mer 23 feb 2022 alle ore 10:04 Lorenz Buehmann
<[email protected]> ha scritto:
I checked Openllet, it has been set to Java 17 - you can
change it in
the POM file:
- set maven-compiler-plugin source and target entry to 11
- change maven-enforcer-plugin Java rule to
<version>[10,15)</version>
Note, the fork is currently set to Jena 4.2.X, not sure if it
will
have
conflicts when you use it with latest Jena
On 23.02.22 08:02, Lorenz Buehmann wrote:
> Hi,
>
> follow up from your Stackoverflow thread, the Jena built-in
reasoners
> do not support SWRL rules - what exactly is supported by
which OWL
> reasoner is documented here:
> https://jena.apache.org/documentation/inference/#owl
>
> None of them is a full OWL DL reasoner, that's only covered
via
> Pellet. And Pellet does also support SWRL. I suggested to use
a
Pellet
> fork like Openllet because the official Pellet reasoner is
still on
> Jena 2.x/3.x and any further version of Pellet is closed
source
being
> integrated in Stardog triple store.
>
> Openllet does support Jena 4.x so in theory it should work. It
would
> be helpful to show your Java/Maven issues, otherwise it's a
wild
guess.
>
> The other option I suggested was to use the Jena rules
instead of
> SWRL. I understand though that if you want to stick to W3C
standards
> (technically SWRL isn't) this won't be an option for you.
>
> Note, the whole reasoner will happen mostly in-memory - don't
expect
> OWL DL reasoning to scale in a large dataset in a triple
store-
that's
> why people designed profiles like OWL RL which can be easily
mapped to
> rule based inference and don't need a tableau algorithm or the
like.
>
> Cheers,
>
> Lorenz
>
> On 22.02.22 18:36, Luca Turchet wrote:
>> Dear list members,
>> I am trying to integrate a reasoner in the Fuseki server. I
first tried
>> openllet but there are some technical issues with java and
maven which
>> currently prevent the openllet installation on a mac.
>>
>> So I tried to launch fuseki with the --conf option using
one of
the
>> reasoners listed at the bottom of this page:
>>
https://jena.apache.org/documentation/fuseki2/fuseki-configuration.html
>>
>> in particular, I tried the examples
>>
>> - config-inference-1.ttl
>>
<
https://github.com/apache/jena/blob/main/jena-fuseki2/examples/config-inference-1.ttl
>> - config-inference-2.ttl
>>
<
https://github.com/apache/jena/blob/main/jena-fuseki2/examples/config-inference-2.ttl
>>
>> which are provided in the documentation:
>>
https://github.com/apache/jena/tree/main/jena-fuseki2/examples
>>
>> When performing the query which should return the result of
an
>> inference I
>> don't get the expected result, like if the reasoner was not
integrated.
>> I am sure that the triplestore I am using is correct and
contains the
>> rule
>> as I have tested it in Protegè using the Snap SPARQL query
tab
with the
>> Pellet reasoner activated.
>>
>> What am I doing wrong? I launch the server with
>> ./fuseki-server
>> --conf=/Users/luca/semanticweb/prova/config-inference-1.ttl
>>
>> Thanks in advance
>>
>> Best wishes
>>
>> Luca
>>
>>
-----------------------------------------------------------------------------------------
>>
>>
>> *Luca Turchet*
>> Associate Professor
>> Head of the Creative, Intelligent & Multisensory Interactions
Laboratory
>> *https://www.cimil.disi.unitn.it/*
<https://www.cimil.disi.unitn.it/>
>>
>> Department of Information Engineering and Computer Science
>> University of Trento
>> Via Sommarive 9 - 38123 Trento - Italy
>>
>> E-mail: [email protected]
>> Tel: +39 0461 283792
>>
