Version 6 (modified by onardmejino, 11 years ago) (diff)


Meeting minutes from 8/15/2007


  • From UW - Jim Brinkley, Dan Cook, Linda Shapiro, Dan Suciu, John Gennari, Onard Mejino, Todd Detwiler, Max Neal, Gary Yngve, Marianne Shaw;
  • From Stanford - Daniel Rubin


  1. Marianne reported that she met with Jim, Dan S, and Todd to discuss possible directions to go in. They

include 1) speeding up existing queries Jena/SPARQL by improving the SQL queries (by preprocessing SPARQL, adding tables/stored procedures, modifying Jena), 2) building a true view language, and 3) extending SPARQL (add path regular expressions, possibly needing a recursive extension). While Marianne has generated the transitive closure of a bunch of properties over the FMA and accessed them through the SPARQL interface, she is working on speeding this up as it is quite slow.

  1. Onard described some of the basic rules that he used in extracting a view of anatomy for RadLex from the FMA. Attached below is a brief description of the steps that were manually carried out to extract that view. This exercise provides the template and the training set for the methods that are currently being developed for the automated view generation tool.
  1. Joshua and Onard are working on a demo that would show AnnoteImage accessing the FMA-RadLex. The FMA-RadLex is an application view from the FMA but the view generated from AnnoteImage is an application view from the FMA-RadLex. See mock-ups below.