RE: Comments for the RDB2RDF Survey Report (as for 20090728)

Dear Satya,

Thank you for your answer. I would like to thank also to all of the authors
to make this survey report. I find it very useful.

 

 

> 1) In the comparison table (Query Implementation part),...In reality, R2O
only works in the second way, it always converts the SPARQL into SQL first.

The paper we reviewed [1],  described the use of SQL "...The main steps of
its executions are: Query & R2O parsing, SQL generation, SGBD execution
result...." and also in a previous publication [2]. 

Yes, according to it, that means there’s no direct execution to RDF
repository and

 

> 2) In fact, my previous point is not true. Rather than using SPARQL,
R2O(+ODEMapster) is using its own query, which is called "ODEMQL". 

We will add this new information to the survey wiki, it would be great if
you could point to a reference/documentation.

We will remove the "both" value from the comparison table for now in the
survey wiki.

The reference is phd thesis of Jesus Barassa  which is in Spanish. [3]

 

> 3) There's a column in the table stating "Number of databases". ...

The column "Number of Datasets" is with regard to top-level column "Data
Integration" and describes the total number of datasets that were integrated
in the given project. For example, DartGrid integrated 70 TCM datasets .

My mistake, I was thinking about “number of databases” rather than “number
of datasets”. So I assume that the column of Data Integration doesn’t
consider where the data  is being distributed into several databases and the
need to integrate the result of the query from multiple databases.

 

 

Best Regards

Freddy

 

 

[3] J. Barrasa. Modelo para la definición automática de correspondencias
semánticas entre ontologías y modelos, PhD thesis. Facultad de Informativa,
Universidad Politecnica de Madrid.

Madrid, Spain. March 2007. 

 

De: Satya Sahoo [mailto:sahoo.2@wright.edu] 
Enviado el: miércoles, 29 de julio de 2009 0:50
Para: Freddy Priyatna
CC: public-xg-rdb2rdf
Asunto: Re: Comments for the RDB2RDF Survey Report (as for 20090728)

 

Hi Freddy,

Thank you for pointing out the issue in the report. 

 

> 1) In the comparison table (Query Implementation part),...In reality, R2O
only works in the second way, it always converts the SPARQL into SQL first.

The paper we reviewed [1],  described the use of SQL "...The main steps of
its executions are: Query & R2O parsing, SQL generation, SGBD execution
result...." and also in a previous publication [2]. 

 

> 2) In fact, my previous point is not true. Rather than using SPARQL,
R2O(+ODEMapster) is using its own query, which is called "ODEMQL". 

We will add this new information to the survey wiki, it would be great if
you could point to a reference/documentation.

 

We will remove the "both" value from the comparison table for now in the
survey wiki.

 

> 3) There's a column in the table stating "Number of databases". ...

The column "Number of Datasets" is with regard to top-level column "Data
Integration" and describes the total number of datasets that were integrated
in the given project. For example, DartGrid integrated 70 TCM datasets .

 

Cheers,

Satya

 

[1] "Upgrading relational legacy data to the semantic web", Barrasa,J.,
Gómez-Pérez, A.,. In Proc. of 15th international conference on World Wide
Web Conference (WWW 2006), pages 1069-1070

[2] "R2O, an extensible and semantically based database-to-ontology mapping
language", Barrasaet al. (2004) In: Second Workshop on Semantic Web and
Databases (SWDB 2004), August, 2004, Toronto (Canada).

----- Original Message -----
From: Freddy Priyatna <fpriyatna@delicias.dia.fi.upm.es>
Date: Tuesday, July 28, 2009 5:44 pm
Subject: Comments for the RDB2RDF Survey Report (as for 20090728)
To: public-xg-rdb2rdf <public-xg-rdb2rdf@w3.org>

> Dear All, 

 

> My name is Freddy Priyatna, currently working in my master thesis
integrating Distributed Query Processor(DQP) with R2O+ODEMapster (one of the
tools mentioned in the Survey Report) in order to make it works with
multiple databases.

> 

> I've been reviewing the survey report (this version :
http://www.w3.org/2005/Incubator/rdb2rdf/RDB2RDF_SurveyReport_01082009.pdf)
and there are some comments that I would like to make :

> 1) In the comparison table (Query Implementation part), it mentions that
R2O works in "Both", which means that the SPARQL query can be executed
directly to the RDF repository or converted first to SQL then executed to
RDB. In reality, R2O only works in the second way, it always converts the
SPARQL into SQL first.

> 2) In fact, my previous point is not true. Rather than using SPARQL,
R2O(+ODEMapster) is using its own query, which is called "ODEMQL". We are in
the progress to enable R2O to support SPARQL.

> 3) There's a column in the table stating "Number of databases". I can't
find detail information about this column. Does "Multiple" value here mean
that the tool allow mapping from multiple databases? How about the query
processing? Does it support DQP as well?

> 

> Thank you

> Best Regards

> Freddy