Provenance Incubator Group Teleconference -- 11 Dec 2009

<trackbot> Date: 11 December 2009

<scribe> Meeting: prov-xg

<scribe> Scribe: Simon Miles

<scribe> ScribeNick: smiles

<lkagal> Great Ivan ! In March, I won't even need a fridge to keep it cold ;)

<ivan> :-)

Yolanda: main topic today, review curated use cases and plan for applying to rest of use cases

<Paolo> http://www.w3.org/2005/Incubator/prov/wiki/Talk:Use_Case_Template

Paolo: been writing up discussion on template on Wiki discussion page, will send link when written up
... Points of curation (1) Provenance is means to end, so use case should describe without ref to provenance
... provenance is then referred to as solution in Problems and Limitations section
... (2) use case should be general, represent set of instances in different domains (e.g. timeliness does not just apply to traffic data)
... also should not be tied to technology, e.g. removed LinkedData from timeliness scenario, put as solution in Problems and Limitations
... (3) should highlight technical challenges of each use case, categorise challenge as about management, use, content
... (talking about http://www.w3.org/2005/Incubator/prov/wiki/Use_Case_Linked_Data_Timeliness)
... (4) separate content, management, usage information
... dicsussion on modifying template, Satya made suggestions to mailing list to remove repititions
... conclusion of discussion: Current Practice is part of Background, but Goal is separate summary from Use Case Scenario
... please put discussions on use case, template pages; Paolo, Satya, Simon will respond

Yolanda: 4 guidlines came from group (last telecon)

Paolo: also added Provenance Dimensions section to template, to specify Primary dimension and some Secondary dimensions

Yolanda: use cases general (not just for one domain): add narrative about technology, e.g. this case applies to LinkedData but would also apply to eScience (refer to other areas not in original case)

<pgroth> +q

Paolo: domain gives reason why use case is there, what inspiration
... can generalise, be clear not intrinsic to domain

Yolanda: where to put this in template?

<pgroth> unanticipated uses? -> other uses?

Paolo: suggests Background, best not to add more sections
... please make use of references (links to Mendeley biblio) in Existing Work section

Paul: already seeing repitition in use case types (e.g. collections)
... could use Unanticipated Uses section for other applications

Yolanda: but in this case is anticipated, and matches existing general use case

<YolandaG> sorry my phone got disconnected, i'll call right back

Paolo: no use case used all sections, Unanticipated Uses not yet used
... was ambiguous whether Goal was user goal or use of provenance
... we are trying to be clear that Goal is user goal then how technology applies

<Paolo> http://www.w3.org/2005/Incubator/prov/wiki/Talk:Use_Case_Template

Yolanda: see, add discussion at link above (but not complete yet)
... new use cases added, more intended but not added yet

<Deborah> +1 for me for intending to add use cases but did not get to it.

Yolanda: set deadline of end of year to have all use cases following curation guidelines

<Paolo> yes

<jmccuske> I've added two use cases but haven't linked to them yet.

Yolanda: owners improve their own use cases, curation team then check

<jmccuske> search in the wiki for "Jim McCusker"/

Yolanda: use cases ready for checking Dec 31, curation team look at for telecon on Jan 8

<pgroth> +q

Paul: is deadline for submission of use cases, or for also complete curation?

Yolanda: deadline (Dec 31) is for submission and self-curation

<Paolo> ok, deadline is fine if we are talking about self-curation

<crunnega> +q

<ssahoo2> I agree with Yolanda

Yolanda: when you believe you have completed your use case, send to one of the three curators
... in Jan, need to move on to how to technically solve use cases

<ssahoo2> I think thats a good plan Paolo

me too

Paolo: instead, send complete use case to all 3 curators, so one available can take it up

Yolanda: please declare if you are intending to add a use case
... I want to add a case where attribution of info is important (someone said it, a journal recorded it...)
... I want to add a case where adding axiom to document, what sources were consulted
... I want to add a case on the provenance of a decision, decide sources, why whichever trusted

<Paolo> all these use cases are extremely important IMO

Deborah: I will add a case on text analytics: text extractors generate sentences, from wide variety of sources
... I may add a case on improving decision processes in producing end conclusions

<Deborah> yolanda will send me a pointer from biology for my text analytics use case. my prior work will cite some intelligence community work

Irini: working on implicit information in RDF, provenance of SPARQL query

Yolanda: if existing use case applicable to you, contact owner to apply your technology to it

<lkagal> +1 for private data use case

Luc: added two use cases: use of private data, prov of decision making (drafts, now to be self-curated)

Paulo: ideas on human level provenance, will connect to Yolanda's use case

<jmccuske> +1 for private data use case, including de-identifying/anonymizing portions of a provenance graph.

<Paolo> not me...

<Deborah> i may join the provenance of decision making as well

Paolo: will go over use cases added a while ago to self-curate

<crunnega> microphone issues

<YolandaG> Christine: you can make any comments on IRC directly, sorry about the microphone

James C: had workshop in April on provenance use cases, will try to write up some, particularly evidence in public policy, engineering/design

<crunnega> Thank you. I would like to review and add to current use cases and then see if I can identify any missing use cases mindful of the deadline.

<Deborah> yes - I am happy to collaborate on that.

Yolanda: discussed with Deborah about why KB/reasoner would produce a given answer, could discuss with James to abstract issues (also maybe relevant to Irini)

Jim Mc: added two use cases on closures from multiple prov graphs, relation between two experiment graphs to determine whether one reproducing other: need to be looked over

Yolanda: mark use case as Draft even if not finished, for others to see
... second case above may be same as Simon's Result Differences case, possibly merge

<lkagal> Sorry, can't unmute am in a noisy environment.

Olaf: not planning to add any new use cases, but will adjust existing

Paul: added two cultural heritage cases, will add Creative Commons case and one on scientific publishing

Yolanda: have been in contact with CC people, so send to her

<lkagal> I'd like to contribute to the private data use case (maybe generalize to any data) and the linked data use cases.

Simon: will look at where dimensions not covered, then conttribute cases if appropriate

Yolanda: will also check dimensions at end of month, discuss on mailing list

Satya: not intending to add more use cases

Yolanda: asks Satya to write up or ask originators to write up use case on prov of scientific hypotheses from

<Aleksey> I would like to contribute to the data quality-aware cases (Information Quality Assessment for Linked Data)

Andreas: intend case on licensing, copyrighting, contracting

Paulo: working on combining proofs, may come up with inconsistent graphs because of different justifications of axioms: please contact him if interested in conflicts in graphs, will coordinate writing case with Deborah

Yolanda: will send suggested F2F meeting dates to list
... telecon next week, but not next two coz holidays

Irini: will scribe next meeting

<ssahoo2> quit

trackbot, end telcon

- DRAFT -

Provenance Incubator Group Teleconference

11 Dec 2009

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output