[OpenID] cryptographics web of trust

[Peter Williams]

Now where can I find a library of sparql queries that deals with lots of
typical "ragged queries", that really showcase dealing with joins
between unstructured semi-formal data sets?

--------------

I've built out my OpenID server-side and consumer-side infrastructure
for querying "ragged data", from several FOAF files: the servers, the
consumer's, the user's, the user's friends, the user's friends of
friends.

This nicely applied wots to control key distribution in forming OpenID
associations, and allows an OP completing AX queries to satisfy the
"informed control" objectives of OpenID Auth.

http://thefigtrees.net/lee/sw/sparql-faq helped me alot. It helped flesh
out AX, assuming it's deployed in a semi-managed UCI environment.

Now where can I find a library of sparql queries that deals with lots of
typical "ragged queries", that really showcase dealing with joins
between unstructured semi-formal data sets?

Why?

1. Many of the AX queries we will be wanting to do in US realty will
address the (property) listing data, as much as it will be addressing
membership and control data used in webSSO. 

Today, the datasets of listings are in 400+ vocabularies. Attempts to
harmonize the vocabulary in the web1.0-era have largely failed, for
socio-political reasons. Yet, cross-source joins and data mining
searches against authoritative copies of listing data are now more
required than ever. My thought is that when supplemented with ragged
searching OpenID AX seems ideal, once OpenID Auth has discovered and
identified the authoritative source(s) of listing data.

2. AX attributes need not be limited to LDIF-like strings. I'm assuming
they can also denote operations the intended recipient MAY perform on a
data source...from generic CRUD operations to (US) realty's URL-encoded
DMQL/RQL-queries. Presumably, one could code URL-encoded SPARQL queries,
too.