Mapping data from RDB to Stardog (OWL)

[Agustín Martínez]

Hi there, i need to populate an OWL2 ontology (a Stardog database) with data from a MySQL database. What tool do you suggest to do this? D2RQ and D2RQ Mapper? any other?

Problem: Insert data (from large tables of health ministry systems) in Stardog to infer some knowledge with a specific ontology (Your reasoner rocks). I'm looking for a tool with extended support to generate constant migration (I need scalability). Perhaps i also need a Excel & XML to OWL tool.

Regards,

Agustín

[Zachary Whitley]

I'd suggest that you use Stardog :)

Stardog 4.1.1 includes an database mapping engine that supports R2RML. This is available with the Enterprise or 30-evaluation license. Excel is easily exported to csv and can be mapped for import to Stardog. For XML you would need an external tool like http://rml.io

If you require a free solution D2RQ would probably work but wouldn't be as nicely integrated into Stardog and it does not support the R2RML standard. To apply reasoning to your data you would either need to use SERVICE queries or import the data into Stardog first.

--
-- --
You received this message because you are subscribed to the C&P "Stardog" group.
To post to this group, send email to sta...@clarkparsia.com
To unsubscribe from this group, send email to
stardog+u...@clarkparsia.com
For more options, visit this group at
http://groups.google.com/a/clarkparsia.com/group/stardog?hl=en
---
You received this message because you are subscribed to the Google Groups "Stardog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to stardog+u...@clarkparsia.com.

[Evren Sirin]

I agree with Zachary :) See this section in the documentation for more details:

http://docs.stardog.com/#_virtual_graphs

There is also an example for importing CSV files here:

https://github.com/Complexible/stardog-examples/tree/develop/examples/cli/virtual/csv

The 4.1.2 release scheduled for next week will have significant
performance improvements for importing large virtual graphs into
Stardog.

Best,
Evren

[justforthe...@gmail.com]

Agustin, You might need an OBDQ solution.

In the past I have tried some generic solutions that didn't give me the nice "fine-tuning" I needed so I opted for a Direct mapping (f.ex you create a Java class that makes the export to Stardog every X time). It is easy to make. This is not too scalable but works fine if the data do not changed too often. Also, it depends on how Stardog is used and how often you can query the prime source of data.

I agree that Enterprise Stardog will be the best solution if your payee is the Ministry or if it can be afforded.

[Agustín Martínez]

Zachary, Evren and JFTH. Thank you all for your answers. I also tried OBDQ for some prototipes. Anyone here has tried Ontomining? Or some data mining with an ontology as input?.

Best, 

Agustin

[Zachary Whitley]

I'm not familiar with OBDQ. What is it? A Google search didn't turn up much.  

--

[Agustín Martínez]

I assume that our friend was talking about "Direct mapping".

[justforthe...@gmail.com]

Sorry, my typo, I meant OBQA (Ontology-Based Query Answering). There is some information on the Web but I can provide pointers if needed.

Alternatively, have you checked Ontop, MASTRO or REQUIEM?