IW Meeting 2013-02-21

From Inference Web

Jump to: navigation, search

IW Meeting 2013-02-21


Contents

Meeting Information

Agenda

Attendees

  • Tim
  • James
  • Deborah
  • Paulo
  • Cynthia

Patrice

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

PROV-WG PR staging document - I'm dragging out their critial path RSN will be modeling provenace of visual strategy invocations (IEEE VIS paper due March) I owe response to Paulo's spreadsheet review. Buried in "stuff" right now :-/ Will not be able to call in next Thursday. I owe response to Patrice for his question at my TWED I owe response to Patrice on WoW workshop paper (today). Tim owes the TURTLE version of Patrice's CMAP, which Patrice updated on github. https://github.com/apseyed/SemantEco-CSV2RDF-extension/tree/master/doc

this is a task in hackpad This is another task in hackpad, take 2.

Jim

  • ICBO post-WWW submission.

proposal - dec 1 target need to think about dataset provenance for ssiii

James

  • Presentation to CogSci Department on OLAP provenance dimension generation and usage.
  • Discussion with BAE on use of this work in Phase 2 ARBITER system.

Cynthia

Preparation for FUSE2 DONE cynthia would like paulo's document from last week - just sent to iw list

Paulo

Wants http://inference-web.org/wiki/PML_3.0_tools to be filled out (it includes link to https://dvcs.w3.org/hg/prov/raw-file/default/reports/prov-implementations.html )

Patrice

Helped out with aggregation paper for ICBO. (vet+qua indiv slant) Starting a paper for Web Observatory Workshop (WoW) (http://wow.oerc.ox.ac.uk/) based on SemantEco + provenance bits Tim and I have been discussing the last couple weeks include a model for 1) graphic based entity 2) the entity it depicts 3) the entity that entity is a specialization of. re-situated the cmap we have for this. (An ISWC paper can grow out of this in connection to health effect modeling w.r.t. chemical (toxicity) measurements and research studies utilizing those data (to support various kinds of search).) Pondering EPA TRI (Toxic Release Inventory) Data Challenge (http://www.epa.gov/tri/university/) and provenence requirements.

Deborah

  • will be away next week
  • discussing with Patrice a submission to the web observatory workshop at www with the provenance slant
  • thanks to jim mcc for getting the icbo paper in
  • thanks to jim and tim for getting the dils paper in
  • fuse - i think james and cynthia might be able to give us some high level discussion of some driving needs for provenance from fuse phase 2 in probably 2 weeks. this will also be related i think to explanation around aggregation and time series.
  • we may be able to generalize from fuse, melagrid, hhs

return to ISWC plans

Discussions

Paulo - can we settle dcterms:FileFormat? https://github.com/timrdf/pml/issues/13 What reservations? We could have a better characterization between Format and Langauge could help us ingest entities in process of explaining provenance of the entities. e.g. pdf with encoding of English. http://dublincore.org/documents/dcmi-terms/#elements-language


<https://github.com/apseyed/SemantEco-CSV2RDF-extension/raw/master/doc/semantEcoProvCmap-2-21-2013.cmap>

 a pml:Document;
   dcterms:format   <cmap>;
   dcterms:language <English>;
.

<cmap>    a dcterms:FileFormat .
<English> a dcterms:Language .

When we find a concrete example of Paulo's reserviations about the relationship between Format and Language, we'll revisit and model it using the terms we're reusing now.

Paulo's claim for a definition for 'format'
---------------------------------------------------

Format: A language is used to determine which tools can ingest a given entity. (pml:Document?)

We are missing the framework for describing the relationship between these "things" format and language. (it's co-occurrence!)

Tim: "... for an application to consume the file. -Paulo"

what are we thinking of putting into iswc? Tim leading a paper - Title: "X" (topic: Using DataFAQs to compare Prizms' output to all other LOD Cloud bubbles.) Patrice will lead one Deborah and jim mcc need to talk

More review of the PML 2 -> PML 3 mappings https://docs.google.com/spreadsheet/ccc?key=0ArTeDpS4-nUDdFdBYkRUQ0hvUWFpTy1ObjlZSTJEaXc#gid=0

-Should we subproperty prov:wasDerivedFrom for pml? Answer: prov:wasGeneratedBy o pml:hadAntecedent => prov:wasDerivedFrom (in ontology already)

The domains/ranges of the property chain above:

E -> Activity -> E implies E -> Entity

pml:Answer subclass of pml:Information

'antecedent' property does not have a range restriction and inherits Entity range restriction from superproperty 'used'. Discussed maybe restricting this further to Information entities, but Jim said he see the range of 'antecedent' as being more open that Information. Hence, we don't currently have a notion of wasDerivedFrom restricted to Information entities (Entity/Entity).


TODO everyone review 7 Feb diagram TODO: Tim to resend the diagram TODO Cynthia to write down explanation for those red links TODO: Tim read Cynthia's explanation and propose PROV modeling.

Facts about IW Meeting 2013-02-21RDF feed
Date21 February 2013  +
Personal tools
Navigation