wiki:minutes_01-08-08
Last modified 10 years ago Last modified on 01/14/08 10:31:56

Meeting minutes from 01/08/2008

Attendance:

  • From UW - Jim Brinkley, Dan Suciu, Onard Mejino, Josh Franklin, Gary Yngve, Marianne Shaw, Jiun-Hung Chen;
  • From Stanford - Daniel Rubin

Summary of Marianne's report:

Nested queries are working; they were implemented as an addition to ARQ. Currently, there are no query optimizations taking place; the subqueries are evaluated and their results stored in memory. A sample query is:

PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>

PREFIX dc: <http://purl.org/dc/elements/1.1/>

PREFIX : <.>

SELECT * FROM <jojo> [CONSTRUCT { ?s dc:wookie ?o }

FROM <bobo> [ CONSTRUCT { ?a dc:booktitle ?c }

FROM <ds-ng-1.ttl>

WHERE { ?a ?b ?c } ]

WHERE { ?s ?p ?o }

]

WHERE { ?e ?f ?g }

Next to be implemented are in-memory recursion and Skolem functions. Further along we need to add recursive support (and possibly a second db format) at the database level.

I am trying to write Onard's RADLEX manipulations as a query to see what other functionality we need.

Also discussed during the meeting was the path that should be taken. Should we try to get Gary's visualization tools to generate queries using SPARQL so that we have a complete demo? Without optimizations we probably won't be able to do this for the FMA, but we could probably do it for RADLEX. Or should we work on views, view composition, and optimizations?