Using Ontop with RDF Data Cube

[Joshua Bezaleel]

Hi everyone,

I am currently involved in a project trying to combine OBDA and OLAP (Cube). The final objective is to be able to execute an OLAP query (slicing, dicing, roll-up, drill-down) to an OLTP (transactional) data source without it having the to undergo a transformation/ETL process (ETL process unluckily will yield another physical cube data, while we do not want to make any other space-taking data/file). That is why we plan to approach this by using ontology as an abstraction for its OLAP features (dimension, measure, slice, etc.) from the data source, in this case, is RDF Data Cube Vocabulary while still keeping it virtual (no physical data cube) as virtual RDF graph by using OBDA tool which is Ontop.

1. Has anyone here been working on RDF Data Cube and Ontop? I tried to access the link mentioning RDF Data Cube at the Ontop's GitHub wiki page but sadly it is already deprecated.
2. Do you think that OBDA and RDF Data Cube approach is the right approach to solve that particular problem?
   And is it possible to use Ontop and RDF Data Cube vocabulary in such way?
3. I already came across this and this topic discussing Ontop and RDF Data Cube but I am still kind of clueless on how to use RDF Data Cube with Ontop, about which part from the RDF Data Cube should be which part on the Ontop (ontology file, mapping file, instances on database) when we are going to "transform" the RDF Data Cube into OBDA using Ontop.
   For example, when we tried to use Ontop, which one from an RDF Data Cube Turtle file (let's say the StatsWales data cube from the W3C) should be a class, which one should be a data/object property, which one should be stored in the relational database that will later be mapped to the concept in the ontology (becoming virtual RDF graphs) using Ontop, etc.? If I get it right, the only thing that should be stored in the relational database and then later mapped to the concept in the ontology is the Observation component?

I am really sorry for the inconvenience because I find the lack of examples and resources to be really hard.

Any kind of help is highly appreciated. Thank you very much.

Best regards,

Joshua

[Guohui Xiao]

Hi Joshua,

We are not aware if someone uses Ontop to deal with RDF Data cube through *virtual graph to support OLAP features*.

1. You can map a database to a RDF Data cube with Ontop. This means that your (virtual) RDF graph is using the vocabulary (i.e., classes and properties) in the https://www.w3.org/TR/vocab-data-cube/. It should be fairly easy to create mappings for leaf concepts (e.g. Observation) using Ontop.  The link you mentioned in your post was creating this kind of mapping. However, they choose to materialize the triples as opposed to keep them virtual.

2.  For OLAP queries (slicing, dicing, roll-up, drill-down), there is no native support from Ontop. At the syntax level, these operators should be expressed in the SPARQL query language but SPARQL does not have these features. So we don't even have a mean to express them in Ontop. 

3. Now at the conceptual level, I am not sure a full virtual approach is appropriate for OLAP style queries since they might be computational expensive. There is a trade-off between space (in the classical data cube) and time (in the virtual approach). A reasonable way to go might be a partial materialisation approach, but this requires further research in both theory and implementation.

Regards,

Guohui 

--
Please follow our guidlines on how to report a bug https://github.com/ontop/ontop/wiki/BugReport
---
You received this message because you are subscribed to the Google Groups "ontop4obda" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontop4obda+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Guohui Xiao, PhD
Assistant Professor with a fixed-term contract
KRDB - Faculty of Computer Science        
Free University of Bozen-Bolzano
Piazza Domenicani, 3                
I-39100 Bolzano, Italy    

http://www.ghxiao.org

According to the Regulation EU 2016/679, you are hereby informed that this message contains confidential information that is intended only for the use of the addressee. If you are not the addressee, and have received this message by mistake, please delete it and immediately notify us. In any case you may not copy or disseminate this message to anyone. Thank you.