IW Meeting 2013-01-31

From Inference Web

Jump to: navigation, search

Contents

Meeting Information

Agenda

Attendees

  • Tim (apologies for missing yesterday's)

Meeting Preparation

Around the room

 * Add a section for yourself 2 hours before meeting.  * Mark any discussion point that you would like to raise during meeting (with DURING MEETING).   * Otherwise, assume that others will read the rest before meeting.   * Also, please be considerate and read others' discussion points before the meeting starts.

Tim

  • I began a paper for IEEE VIS due mid-March, "Content-Preserving Visual Artifacts" that will demonstrate some advantages of including provenance in the visualization process, visualizing linked data, and encoding the provenance using linked data.
    • This will motivate some development of PML 3.0
    • I reviewed and refactored my age-old RDF visualization system, and there's a lot of provenance already there that I need to review and tweak. Again, this is more motivation to develop PML 3.0.
  • After the VIS paper, I'm turning to write an ISWC paper to compare Prizms to the rest of the LOD Cloud. This will be intensive until Jun, but will not be heavy on provenance.
  • The follow-on to VIS would be to enable question answering about the visualizations, which would elucidate further benefits of capturing provenance and using linked data for both the data  and provenance. LOTS of motivation for PML 3.0 designing here, but this task is 3rd on my list of papers and can't receive a significant portion of my attention until late summer, early fall.
  • The follow-on for the ISWC paper is demonstrating general visualization techniques for arbitrary RDF graphs (e.g. SPO Balance and Between The Edges), and pairing that up with the initial stages of visual strategy development (selecting, filtering, and deriving the domain data for a visualization) that uses existing linked data and SADI. This would line up to feed into the relatively blank portion in the VIS paper by providing a declarative representation of the visual strategy AND the provenance of its creation (via user interaction).
  • By then, I expect that I'd be done with the core of my thesis, with the only thing left potentially being an overall evaluation of the already-evaluated sub-pieces.
  • Nick and Hugo are meeting with me at 4pm.
  • Reminder that I announced the hasMetaBinding grep at http://inference-web.org/wiki/IW_Meeting_2013-01-24#Discussions
  • Reminder that I created the 3 requested examples at http://inference-web.org/wiki/IW_Meeting_2013-01-24#Discussions
  • apologies for missing yesterday's meeting.

Jim

I see 2 papers going forward:

SIO/PROV mapping which shows value in these ways:

  • addresses identity of biological entities issues from OWLed 2010 paper
  • Enables a model for expressing cancer biological specimens and genetic variants that allows for effective analysis and visualization. An example visual is at (ask Jim)

Follow-up to aggregation paper that uses the model from the SIO/PROV paper as an example to demonstrate the value of aggregation and identity semantics.

James

  • Continued work on datacube provenance dimension paper.
  • Discussions with BAE on approval for publication.

Cynthia

  • Review various documents, PML3, PROV-O, Fuse 2 indicators, ARBITERAAssumptions.
  • Start FUSE-Datacube discussion wiki page.

Paulo

Patrice

Deborah

Discussions

Pre-meeting

Looking for Q&A (not healthdata)

  • Semanteco
  • melagrid
  • Tim's VIS
  • Global warming report

done: Tim to tset up a space on https://hackpad.com/ TODO: everybody make an account.

http://inference-web.org/wiki/PML_3.0 links to:


james' assumption 6 / requirement ASSUMPTION 6: Document trace-back functionality.   RPI will be able to trace back to specific source documents (and fragments of source documents) used by ARBITER to generate answers for emergence-related questions.


Using prov:atLocation as the superclass of a qualification of pml:atIndex. The 


done: Cynthia make an example that shows 2 query/answer chains, and how the elements of each chain is grouped into the chain. (listed in http://iwepad.titanpad.com/160 as http://inference-web.org/proofs/tonys/tonys.owl and others)


http://www.w3.org/TR/prov-o/#inverse-names-table

Cynthia's notes about deprecation http://inference-web.org/wiki/Pml2-concept-discussion


done: Tim to grep provweb for the Prettys :-)

captured: Tim to review prov-dictionary w.r.t Map.

Specifying subactivities in plans: 
Class: StartAForestFire
EquivalentClass: dc:hasPart some (Ignition that prov:wasInformedBy some (DropTheMatch that dc:isPartOf this)) 

captured: ^^^ how does this reconcile with Association subclass of Agent? (Tim->Jim) captured: Tim write it up.


counter example to transitivity: http://www.w3.org/TR/prov-constraints/#generation-use-communication-inference


http://www.w3.org/TR/prov-dm/#section-derivation

http://prefix.cc/prov

Facts about IW Meeting 2013-01-31RDF feed
Date31 January 2013  +
Personal tools
Navigation