[dotNetRDF-Support] Consuming "Triple-store" table in a relational database

[Alexander Y. <a.a...@ho...>]

Hello,
 
We are looking for a solution to a recurrent design problem: consumption of (external) open data in an application business logic. The open data set and the application belong to different segments of same domain or even to different domains. The structure and content of the open data set are controlled by an independent organization.
 
The open data set is available in XML format. It is also ReadOnly for us.
 
We want to achieve a level of abstraction and independency from possible changes in the structure and content of the open data source. There is no 100% solution to this of course, so we are aiming at implementation of our application business logic with a certain level of tolerance to the changes in the open data set. It is expected that the implementation of application business logic will remain unchanged  until it can produce a satisfactory output.
 
Another part of the solution is that we want to accumulate data  with our segment of the domain and then query that data in conjunction with the open data set.
 
I was thinking about the following solution:
1. Convert XML to RDF. Serialize RDF in a table an relational database.
2. Use SPARQL to query to query  the RDF table.
3. Convert SPARQL query results to XML or JSON format which then can be consumed in C# or JavaScript program respectively.
 
There are few reasons I am considering this solution. For one, we need integration with RDF that we will generate. Secondly, because it is a legacy WinForms project where we already use SQL Server as a repository of data we do not want to install and administer additional server components like Virtuoso, Sesam etc. We also cannot afford buying additional server components for each customer.
 
Question: Can DotNetRdf consume data from a relational table containing triples? A table with three columns respectively.

 
Thank you!
 
Best regards,
Alexander Yarushin