---- On Mon, 08 Aug 2016 04:20:51 +0200 Adrian Walker
<[email protected]> wrote ----
> Simon,
>
> May I ask please, what is the end-user application that is to be supported
> with the techniques you describe?
It is difficult to describe for me because it is a research project. It is
basically a semantic search engine, which shall understand the semantics behind
data. As an example, if there are multiple organizations, which provide models
for a person, I would like that they (or someone else) provide ontologies for
these models. Once they do, my search engine can index the ontologies and
instances of the person classes. Now, if someone searches for a person I of
course want that it doesn't matter for which ontology they search, the search
engine should find everything that semantically makes sense to be a person.
I need a query engine, which can operate as the interface of the search engine
and I thought that SPARQL would be a good at the beginning. I don't know if
SPARQL is too low level for my use case. I thought that with inference support
I can at least create queries that abstract over what exactly the URI of a
person is and therefore let me search for all kind of persons. Maybe I'm wrong
with that and instead I need another query engine, which internally creates
SPARQL queries that find everything that can be a person.
