IW Meeting 2013-01-17

From Inference Web

Jump to: navigation, search

Contents

Meeting Information

Agenda

Attendees

  • Tim
  • Jim
  • James
  • Paulo
  • Deborah

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-O term occurrence in http://healthdata.tw.rpi.edu/sparql:

--------------------------------------
| term                      | count  |
======================================
| prov:wasAttributedTo      | 316605 |
| prov:Activity             | 85182  |
| prov:wasGeneratedBy       | 84238  |
| prov:hadPlan              | 83628  |
| prov:qualifiedAssociation | 83287  |
| prov:Association          | 77249  |
| prov:used                 | 35902  |
| prov:specializationOf     | 28443  |
| prov:wasDerivedFrom       | 9873   |
| prov:generatedAtTime      | 8688   |
| prov:startedAtTime        | 7664   |
| prov:atLocation           | 7055   |
| prov:alternateOf          | 6633   |
| prov:wasInformedBy        | 6272   |
| prov:endedAtTime          | 836    |
| prov:agent                | 487    |
| prov:wasAssociatedWith    | 487    |
| prov:value                | 314    |
| prov:Plan                 | 10     |
--------------------------------------

Jim

  • Finished up the NCI grant, will now finish up the DILS paper.
  • Suggestion: Submit the SIO/PROV mapping to ICBO?

James

  • Write-up for annual FUSE report. (provides explanation of Arbiter using IWBrowser)
  • Draft for plan composition paper, to share internally at RPI.
  • Case studies for explanation in FUSE Phase 2. - dlm - thx did you include a pointer to these on the password protected site?

Cynthia

  • looking into OLAP
  • looking for OLAP software to use on top of a triple store.
  • OLAP4j (java) targets SQL, but nothing about a triple store.
  • Virtuoso said they can do limited OLAP with GROUP construct.
  • Tim's disheveled running notes on Data Cube: https://github.com/timrdf/csv2rdf4lod-automation/wiki/RDF-Data-Cube
  • JPM Comment: olap4ld implements an XMLA service that relies on RDF Data Cube Vocab.

IWComponents licensing

Paulo

Some funding from ARM for climate model that DOE uses. Create prototype that answers:

  • user of ARM dataset, how to know that my dataset is flagged with low quality?
  • learn that my version of the database has been replaced by a newly calibrated data?
  • what are the flags and what do they mean?

They have a DB with the data and provenance, but it needs to be mapped. To PROV?

Paulo's paper outline:

As per our conversation last week, I have an outline for our PML3 paper in place and in the process of filling out some of its sections. At the moment, I would like to share a draft introduction with you and see your comments/reactions to it. The draft is attached below in the body of the message.

Many thanks, Paulo.

According to W3C, “provenance of a resource is a record that describes entities and processes involved in producing and delivering or otherwise influencing that resource.”

We would like to highlight a couple of interesting aspects from this definition:

a. The record is considered provenance regardless of anyone’s perception that the account is true, incomplete or inconsistent; and b. As a record, provenance is expected to report entities and processes involved with the production and delivering of the resource of concern (RoC) without tailoring its content in a particular way that would facilitate the sharing of its content with a specific agent or community.

Regarding (a), the provenance community recognizes the need for mechanisms to verify whether provenance recordings are sound and based on trustworthy factual data [cite IW-Trust, Missier’s work, Ludaescher’s work]. In this paper we focus on the soundness of the provenance by discussing whether provenance traces can be classified into justifications, and more specifically, when justifications are described under appropriate formalisms to be in conformance with well-establish proof theory notations. Regarding (b), we would like to highlight that the knowledge exchange between provenance producers and provenance consumers is rarely done in a single and direct manner. Instead, we observe that the follow processes are more likely to occur: provenance producers generate provenance records; provenance consumers or provenance consumer proxies read and interpret provenance records; when proxies are involved, they may generate explanations that are the actual knowledge perceived by provenance consumer. Because of this complex knowledge exchange pattern, we believe that effective uses of provenance need to be discussed in the context of explanations, and if possible, within a question-answering communication paradigm.

Patrice

Deborah

  • grant and travel stuff dominated -
    • best paper in health track at hicss .... but challenging trip
    • fuse extension paperwork in
    • almost done NIH proposal update
  • iaai paper going in today on semanteco
    • are there other papers going in place in the next few weeks other than dils with mccusker/lebo? - jim mentions icbo for sio/prov - this may be a return to
  • return to - PML 3
  • return to - licensing issues as needed by a report cynthia has the lead on for fuse - apache 2

Discussions

Facts about IW Meeting 2013-01-17RDF feed
Date17 January 2013  +
Personal tools
Navigation