SIMILE PI phone conference, 22-Jan-04 1100 EDT/1600 BST

SIMILE PI phone conference, 22-Jan-04 1100 EDT/1600 BST
  
866-639-4752 or +1-574-935-6705
PIN: 2536617
     
irc://irc.w3.org:6665/simile

Agenda

1. Round table update from Pis

2. Logistics for arrival of new hires

3. Review project task list - see below


1. ARTSTOR AND OCW DATASETS

Progress update from Mark:

- Now possible transform OCW data just with XSLT, no need to use Perl. The
transform does not do everything the Perl transform did, but this avoids all
the team installing Perl on their machines.

- Artstor topic, geographic and subject fields are now organized
hierarchically where appropriate.

- New ANT automatic build script in CVS allows all team members to rebuild
datasets, avoids need to upload entire datasets to CVS, team members just
need to retrieve updated XSLT scripts from CVS.

- Added type information to OCW dataset.

To do:

- Adopt a common way of displaying names in Artstor and OCW.

- Need to change the way typing is done in OCW so it is more compatible with
the existing LOM schemas.

2. HAYSTACK / SIMILE

Progress update from Steve

Progress update from Andy on Joseki / Haystack integration

To do:

- Need to update simile.ad file so Haystack can display revised OCW dataset.

- Now hierarchical information has been added to datasets, Haystack needs to
process this in faceted browser.

3. CUSTOM BROWSER

Progress update from Mark:

- Can now display both Artstor and OCW data

- Can display hierarchical facets e.g. geographic, subject and topic

- Uploaded to CVS, Rob has been able to retrieve and run on both Linux and
Windows.

To do:

- Write interface so it is possible to switch between using Lucene for
queries and RDQL.

- Add text search boxes to facets when it is not possible to display all
facet values.

- Add paging to facets when it is not possible to display all facet values.

- Fix Ant / XSLT / MTXSLT / Saxon 7 / ENTITY / DOCTYPE bug.

- Need to fix Ant build script so tasks that call XSLT, unzip or Jena
Schemagen only rebuild when necessary.

- Add Jena persistant model support.

- Demonstrate inferencing between the two datasets.

- Display facet frequency.

- Allow both alphabetic sorting of facets and sorting by facet frequency.

4. CVS ACCESS

Progress update from Mark:

- CVS is now better organized, possible for everyone to build datasets, they
just require Ant and Java on their machine, 
All other dependencies in CVS.

- Requested login in for David Karger, need to check if anyone else has CVS
access problems.

5. IDENTIFYING CORPUS SUBSET

Kevin on holiday this week. 

NON-ACTIVE TASKS

- Haystack user testing

- Making datasets available via Joseki

- Custom browser user testing

- Exploring data using Brownsauce or RDFNavigator

- Creating infrastructure at simile.mit.edu

Dr Mark H. Butler
Research Scientist                HP Labs Bristol
mark-h_butler@hp.com
Internet: http://www-uk.hpl.hp.com/people/marbut/

Received on Wednesday, 21 January 2004 15:34:20 UTC