IW Meeting 2011-01-20

From Inference Web

Jump to: navigation, search

Contents

Meeting info

Attendees

  • Tim
  • Li
  • Stephan
  • Cynthia - are you on Skype or telephone? both
  • Geoff

waiting:

  • Paulo (regrets)
  • Deborah (regrets)

Agenda

  • 1 - determination on who at utep is helping take the lead on the layering document with tim. (action item for paulo)
  • 2 - let me know if there is anything anyone would like rpi to raise in voting for the provenance working group. (probably this is mostly also aimed at paulo but feedback from anyone is welcome). comments can be from the reading of the provenance incubator report or the draft charter or both.
  • 3 - tim to bring up what deborah and tim discussed of tim's compilation of pml and analysis
  • 4 - action item review -
  • 5 - status from geoff and cynthia both on work and on whether we are likely to get a near term paper submission in.
  • James is planning a change request to add "accounts".

Pre meeting

Meeting time next semester

Is 11am Thursdays ok for you?

  • Tim - yes
  • Cynthia - yes
  • Paulo - yes
  • Li - yes
  • Deborah - yes
  • Geoff
  • Jim - yes
  • Nick - yes
  • Jitin - yes
  • Stephan - yes

Discussion

Proof combining paper

Will not make the CADE paper, aiming for workshop.

Li wants to follow up to understand equivalence options.

Cynthia working on integration, small possibility to make deadline in 10 days. Geoff waiting on Cynthia to integrate.

Geoff: Rushing to do research to make a paper makes you do bad research.

Layering

What is layering?

Tim: just a partitioning?

Stephan: Renaming in prep for FOO? cutting out unnecessary structure.

Stephan: aligning with other ontologies.

DONE: Tim to dig up OPMO's ontology URL. (It's URL is encoded in PML at http://inference-web.org/w/images/f/f4/Opmo-20101012.owl.pml.ttl)

Management: how to manage ontology dev requests

Tim: Assume we have two layers. What are the factors that put it in the top or bottom? Li: We MUST have explanation for why it is in one layer or another. Identify concrete points:

  • e.g. foo is going to use it
  • use case X of project Y needs it (and so do 5 other use cases from 3 other projects)
  • tool Q uses it to render labels.
  • Concept: class, property, structure, restriction
  • Layer: fundamantal, customized
  • Use-case: a concrete example (text excerption) from Input, explaining requirements to PML ontology
  • Input: project (e.g. SPCDIS, MDSA, DQSS, AeroStat, CISRO collaboration, TPTP, LOGD), tool (e.g. PML API, IW Browser), Prov WG...
  • Request: put a concept CX to layer LX because of use-case UX from input IX, and it may incur impact to input IY in use-case UY.

Li: Start with one example, include more examples from more projects.

Stephan: Document fundamental aspects: What HAS to be there. Start with IW Browser, (e.g., Rec 1 from prov-xg) - http://www.w3.org/2005/Incubator/prov/XGR-prov-20101214/#Recommendations

Stephan: we need a list of the fundamental concepts

Cynthia: Some things are never being used

  • e.g. WebSite never used.
  • How do we know? Cynthia has never seen it or used it.
  • Stephan has not used it.
  • usually they only have Information and hasURL


Li: "When am I done?" (what does done look like?) what is outcome of doing {use case, tool documentation, etc.}? Li suggests: "yes/no" requests that can be answered for each use case?


Roles

Roles are not straighforward

Ordering of antecedents

antecedents are ordered lists - hard to just get the set without walking the list.

Li: naming vs. structure.

un-ordered list: bag order list: seuqence, list other approach:

list rdfs:member member1 list rdfs:member member2 list rdfs:member member3 member1 x:next member2 member2 x:next member3

Stephan: order should not be required at the foundational level?

Should antecedents be ordered?

Inputs

http://tw.rpi.edu/web/project/SPCDIS http://tw.rpi.edu/web/project/MDSA http://tw.rpi.edu/web/project/DQSS http://tw.rpi.edu/web/project/AeroStat

Action items

http://inference-web.org/wiki/PML_2.1_-_a_Layered_Approach

  • Geoff to answer Li's questions (lattice)
  • Tim write up LOGD use case (tie to FOO)
  • Cynthia to put together list of what PML is needed for IW Browser to work. - put materials at/near http://inference-web.org/wiki/IW_Browser
  • Stephan to start a page with the fundamental concept list - figuring out what is fundamental.
  • Cynthia, Stephan, and tim to review tools#tools and PML Datasets pages to heljp peopel find the data.
  • Tim to coordinate with Nick and Jitin.
  • Tim will find good use case examples to base ours off of.
Personal tools
Navigation