Hello,
I just wanted to give you a status update on the progress I have been
making with the development of the VCL broker and get some feedback on
the approach.
1. I have created OWL/RDF ontology for VCL Image. This ontology will
be used by the Cloud Broker tool and can be used by any software agent
that wishes to interact with VCL. The ontology is available on the
public domain at
http://ebiquity.umbc.edu/resource/html/id/343/VCL-Image-ontology.
2. While I was able to successfully integrate the user interface with
the Joseki server reading test VCL Image data; I have hit a roadblock
while integrating it with VCL's MySQL database to read directly from
the IMAGE table. The reason is that Joseki will only read data in RDF
format and not in SQL format. And so I have to write a script to
manually query VCL Database and then create a RDF datastore out of
that data which will be queried by the Broker tool. While this
approach works, it is not a good design for data integrity and will
require me to run this process frequently to synchronize the VCL
database with the Cloud broker database . I want to avoid creating a
new database that will have to be maintained by VCL Admins.
So, now I am changing my approach to use another SPARQL
platform called D2RQ (http://d2rq.org) which claims to allow
accessing relational databases as virtual read-only RDF graphs. I am
hoping I will be able to integrate it with the Cloud broker tool and
remove the additional system process of converting VCL data into RDF
format .
I have also updated this on the JIRA issue for this project. Would
love to hear from anyone who has worked with D2RQ .
--
regards,
Karuna
____________________
Karuna Pande Joshi
PhD Candidate,
CSEE Dept, UMBC
kjos...@umbc.edu, karuna.jo...@umbc.edu
