Re: JENA in TinkerPop?

[Olivier Rossel <ol...@gmail.com>]

I might misunderstand the concept of "property graph" but it sounds
like a superset of RDF graphs.
The most visible feature of property graphs that is missing in RDF
graphs, is properties set on relationships.
Right?

So it sounds possible to use Tinkerpop to traverse/query a RDF graph.

Am I wrong?

On Wed, Sep 6, 2017 at 4:25 PM, Karl Pietrzak <ka...@gmail.com> wrote:
> Thank you so much, Andy, that was very informative (especially the
> presentation).
>
> If I have some OWL datasets and working on causal modeling, it seems like
> I'm more in the "RDF view of the world" vs the "property graph" view of the
> world, which pushes me toward Jena over Tinkerpop, if I'm understanding
> correctly.
>
> What other software would you recommend?
>
> On Wed, Sep 6, 2017 at 5:27 AM, Andy Seaborne <an...@apache.org> wrote:
>
>> Hi Karl,
>>
>> Apache Jena is based on RDF and linked data.
>>
>> The graph data model is different to Tinkerpop's which is property graphs.
>> Neo4j provides property graphs with extensions.
>>
>>
>> RDF - a W3C standard, along with SPARQL, an RDF Query language are not
>> based on property graphs.
>>
>> Property Graphs and RDF come from different heritages; addressing
>> different problems.
>>
>>
>> Data modelling is very important in the RDF view of the world - writing
>> down what the data means. This goes with publishing because it means the
>> data says what the data means to whoever is consuming it. In publishing,
>> you don't know the app/people are reading the data.
>>
>> RDF uses URIs because globally unique names, that you can look up to know
>> what they are naming, are important for publishing data and for data
>> integration of disparate systems,often systems not under the control of the
>> group doing the integration.
>>
>> Property graphs are more aimed at capture and processing rather than
>> modelling and integration.  Currently, especially in analytics (where the
>> $$$ around processing is).  Applications and database are built together
>> for a defined purpose.
>>
>>     Andy
>>
>> https://www.slideshare.net/andyseaborne/two-graph-data-model
>> s-rdf-and-property-graphs
>>
>>
>> On 05/09/17 22:42, Karl Pietrzak wrote:
>>
>>> Hello everyone!
>>>
>>> I'm trying to familiarize myself with JENA / Tinkerpop /etc., and I'm
>>> trying to figure out where JENA hits into the Tinkerpop "ecosystem".
>>>
>>> Would it be a "data system provider
>>> <http://tinkerpop.apache.org/providers.html#data-system-providers>"  like
>>> neo4j because of TDB?
>>>
>>> But it seems like Fuseki
>>> <http://jena.apache.org/documentation/fuseki2/fuseki-run.html> provides
>>> an
>>> HTTP layer on top of Tinkerpop, putting it in competition with Gremlin
>>> Server <http://tinkerpop.apache.org/docs/3.3.0/reference/#gremlin-server
>>> >?
>>>
>>> Any hints would be greatly appreciated.
>>>
>>> Thank you!
>>>
>>>
>>>
>
>
> --
> Karl

Re: JENA in TinkerPop?

[aj...@apache.org]

I think Andy's response from a few years is still a pretty good one.

https://mail-archives.apache.org/mod_mbox/jena-users/201508.mbox/%3C55D700FA.10705@apache.org%3E

ajs6f

Olivier Rossel wrote on 10/11/17 3:20 AM:
> I might misunderstand the concept of "property graph" but it sounds
> like a superset of RDF graphs.
> The most visible feature of property graphs that is missing in RDF
> graphs, is properties set on relationships.
> Right?
>
> So it sounds possible to use Tinkerpop to traverse/query a RDF graph.
>
> Am I wrong?
>
> On Wed, Sep 6, 2017 at 4:25 PM, Karl Pietrzak <ka...@gmail.com> wrote:
>> Thank you so much, Andy, that was very informative (especially the
>> presentation).
>>
>> If I have some OWL datasets and working on causal modeling, it seems like
>> I'm more in the "RDF view of the world" vs the "property graph" view of the
>> world, which pushes me toward Jena over Tinkerpop, if I'm understanding
>> correctly.
>>
>> What other software would you recommend?
>>
>> On Wed, Sep 6, 2017 at 5:27 AM, Andy Seaborne <an...@apache.org> wrote:
>>
>>> Hi Karl,
>>>
>>> Apache Jena is based on RDF and linked data.
>>>
>>> The graph data model is different to Tinkerpop's which is property graphs.
>>> Neo4j provides property graphs with extensions.
>>>
>>>
>>> RDF - a W3C standard, along with SPARQL, an RDF Query language are not
>>> based on property graphs.
>>>
>>> Property Graphs and RDF come from different heritages; addressing
>>> different problems.
>>>
>>>
>>> Data modelling is very important in the RDF view of the world - writing
>>> down what the data means. This goes with publishing because it means the
>>> data says what the data means to whoever is consuming it. In publishing,
>>> you don't know the app/people are reading the data.
>>>
>>> RDF uses URIs because globally unique names, that you can look up to know
>>> what they are naming, are important for publishing data and for data
>>> integration of disparate systems,often systems not under the control of the
>>> group doing the integration.
>>>
>>> Property graphs are more aimed at capture and processing rather than
>>> modelling and integration.  Currently, especially in analytics (where the
>>> $$$ around processing is).  Applications and database are built together
>>> for a defined purpose.
>>>
>>>     Andy
>>>
>>> https://www.slideshare.net/andyseaborne/two-graph-data-model
>>> s-rdf-and-property-graphs
>>>
>>>
>>> On 05/09/17 22:42, Karl Pietrzak wrote:
>>>
>>>> Hello everyone!
>>>>
>>>> I'm trying to familiarize myself with JENA / Tinkerpop /etc., and I'm
>>>> trying to figure out where JENA hits into the Tinkerpop "ecosystem".
>>>>
>>>> Would it be a "data system provider
>>>> <http://tinkerpop.apache.org/providers.html#data-system-providers>"  like
>>>> neo4j because of TDB?
>>>>
>>>> But it seems like Fuseki
>>>> <http://jena.apache.org/documentation/fuseki2/fuseki-run.html> provides
>>>> an
>>>> HTTP layer on top of Tinkerpop, putting it in competition with Gremlin
>>>> Server <http://tinkerpop.apache.org/docs/3.3.0/reference/#gremlin-server
>>>>> ?
>>>>
>>>> Any hints would be greatly appreciated.
>>>>
>>>> Thank you!
>>>>
>>>>
>>>>
>>
>>
>> --
>> Karl