2012Jul18

From FilteredPush
Jump to: navigation, search


Etherpad for meeting notes: http://firuta.huh.harvard.edu:9000/FP-2012Jul18

Agenda

  • Botany 2012 reports
  • Discuss messaging, SPARQLPuSH

Non-Tech

  • Media proposal from Harvard-Smithsonian Center for Astrophysics
  • Collaborations
    • Kepler
    • Specify/Symbiota
    • SCAN TCN
    • NEVP TCN

Reports

  • Paul
  • David
    • Deployed new ClientHelper code and PHPLibraries with Symbiota/Morphbank on local machine (with date validation code)
    • Bug fixes to SPARQLpush, added a SPARQLPushLoader class to the PHP libraries for injecting annotations via Symbiota
    • Modified Symbiota code to use the new ClientHelper php to generate annotations and load them into SPARQLPush via the SPARQLPushLoader (working implementation of FP-Lite)
  • Lei
    • Regarding to the problems discussed in "what's missing or not clear", proposed solution for the mapper/adapter, had discussion with Maureen.
    • Began to refactor the annotation processor/mapper/adapter
  • Jim
    • SCAN: grant paperwork has been completed, grant start date is 1 July 2012, and funds are in hand. Paul and I travel to Tempe next month.

Notes

FilteredPush Team meeting: 2012 July 18

Present: Jim, James, Lei, Bertram, Maureen, Bob, David, Paul

Agenda:

  • Botany 2012 reports

James: Talk went well. Connectivity problem half way through demonstration, had to switch to sceen capture.

User Feedback: Can I filter messages in the way that I manage messages?

User Feedback: Can I get an RSS feed?

Bob: RSS is working now with SPARQL Push.

James: Folks attached to resources. Solanum PBI interested in a scratchpad client to FP (Scratchpads are in Drupal). SilverBiology interested in disussing API and being involved in testing. JSTOR interested in using annotations in GPI. Ed Gilbert happy to deploy when we are ready.

Bob: Shouldn't be much work for David to wrap the client in Drupal.

Jim: Additional resources, how do we want to use funds from SCAN TCN, use for current staff and extending time, adding people?

Bob: We know that we need some attention to UI issues. Might be a diversion of effort for Maureen/Lei to put full focus on implementing UI elements. Probably UI designers who are looking for work now (based on feedback from feeling folks out for proposal).

Discussion: Model train the trainer for Ed. We help deploy Client code into SCAN and NE VP symbiota instances, Ed now in position to deploy client into other symbiota instances.

Bob: Are there differences in requirements for systems in which you need to reason on the annotations and those in which annotations can be returned as static documents. Differences arise when you consider the expectations of users with regards to their interests. If users want to know the list of what annotations meet their interests at the point the annotations were published, this entails a document store. If on the other hand users expect to see a current view of the world, then reasoners can add arbitrary tripples and results may not reflect original annotations.

James: At realm of government, there would be an ability to retrieve exactly what was said by a particular person. A provenance chain would be very important.

Maureen: Both ALA and morphbank treat annotations as documents.

James: In botanical community, they are things on bits of paper, largely new determinations.

Maureen: Cognitive model is yellow sticky.

Paul: ALA, Morphbank, Annotations on herbarium sheets all suggest that large portions of user community will wish to find out what the annotator originally said.

Bob: Provenance gets to the heart of the issue.

James: Do you annotate the consensus, or is the consensus generated from opinions.

Bob: Depends if the conversations are linear.

Paul: It sounds like there is a requirement for users to be able to follow the provenance of annotations in their original form.

Bob: Not necessarily distinguishing triple stores from document stores.

Maureen: Document store likely to be most efficient way to retrieve original provenance.

Bob: Not clear that a document store can solve all the problems on the table. Probably can for ApplePie, not clear in the more general case.

Maureen: As long as you don't try to make a triple store into a document store.

Bob: W3C group embroiled in exactly this question...

Non-Tech

  • Media proposal from Harvard-Smithsonian Center for Astrophysics

Jim: I propose that Paul or I circulate the written proposal that we've received from the Smithsonian media group. I would then appreciate hearing if people endorse this proposal and support paying this expense from the FP grant.

TODO: Paul to circulate, people to send feedback to Jim.

  • Collaborations
    • Kepler

Jim: need to get a draft budget to local grants person right away. About to leave this weekend for two weeks.

TODO: Bertram: communicate with Paul about what items should be included in Harvard's budget. Bertram: Call with Paul tomorrow? James, Paul, Jim available anytime tomorrow.

Bertram: 1:15 Eastern tomorrow. EVO call.

    • Specify/Symbiota
    • SCAN TCN
    • NEVP TCN
    • GBIF collaboration (Nodes meeting next Mon-Tues.)

James: Member of Canadian node, NorthAmerican nodes meeting next week, just got agenda and on agenda to talk about FP...

Bob: Would presentation of video with audio added meet needs?

James: Need to find out more from conveners.

James: Data publication also on agenda. Discussion of second copy of index.

Bob: That's interpretation...

James to follow up with conveners, contact us if more is needed from rest of team.