IW Meeting 2013-07-11

From Inference Web

Jump to: navigation, search


Meeting Information



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.


I'm running a survey of the authors of the 900 datsets tagged "lod" on datahub.io. Survey questions listed at https://github.com/timrdf/lodcloud/wiki/Survey-1-2013-Jun-22 and in google spreadsheet. The survey asks about what products have come from it (papers, applications), and how they produced it. It also asks them to rate how difficult it was to fulfill metadata requirements, and what factors influenced those ratings. It also asks if they want to see another LOD Cloud Diagram.

  • Adopting pml:Question and pml:Answer to model my LOD Cloud survey results.
    • (from dlm - great - are you finding benefit in it?)
    • Attempted to send to authors of all 900 "lod" tagged datasets in datahub.io
    • About 50% dataset response coverage (thanks to some aggregate responses for 244, 53 datasets from two people)
  • Protege is giving syntax errors on http://provenanceweb.org/ns/pml#
    • what is causing that?

I hit a snag trying to parse out tools etc from the dcterms:description. #regexhell Otherwise, this would all be hosted already using Prizms.


   a pml:Answer ;

   dcterms:isReferencedBy <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/version/2013-Jul-07> ;
   void:inDataset <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/version/2013-Jul-07> ;

   # The subject of the survey 
   dcterms:subject <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/the-drug-interaction-knowledge-base> ;  

   # The question that I asked the Participant.
   prov:wasDerivedFrom <http://purl.org/twc/lodcloud/source/us/survey-1-questions/lod_3_tools> ;      

   # Participant's response
   dcterms:description "Python, RDBLib, MySQL, SQLAlchemy, D2R -- several custom scripts are not available puiblicly" ; 

   # parsed from dcterms:description above
   dcterms:subject <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/tool/SQLAlchemy>,     
                            <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/tool/RDBLib> ; 
                            <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/tool/MySQL> ; 
                            <http://purl.org/twc/lodcloud/source/us/dataset/survey-1-results/tool/D2R> ; 

  # Intentionally NOT attributing the answer to the person who provided it.

   a pml:Question;
   skos:altLabel "1) Does the tag "lod" on the dataset %DATASET% mean "Linked Open Data"?";

   a void:Dataset;
   prov:specializationOf <http://datahub.io/datsaet/the-drug-interaction-knowledge-base>







  • At ESIP Federation Summer Meeting this week
    • Some great discussions on the new Prov W3C
    • Presentations and discussions on creating a new Prov-ES extension for earth science
    • Will be having a discussion later today on where to go next with Prov-ES
  • Discussions about software provenance
    • a lot of software tools that do aggregation, subsetting, constraints, server-side functions, etc... but do NOT add any information to the data products (attributes or whatever) that says what data went into the creation of the data product. And the user sometimes doesn't know.
  • Discussion of a tools ontology about software, applications that can do certain things, what kind of software can access data
    • Called ToolMatch
    • Part of the ESIP Semantic Web Cluster (Peter is lead on this cluster)
    • I'm talking with Eric Rozell about what he did with this (he originated much of the work over the last year+


  • health issues dominated the last little while. just starting to re-emerge once we get a schedule set



  • propose to model hasAuthorList with Authorship + atIndex.
    • This is kinda what we do with TW Ontology. hasAuthor -> Author with an index that shows the ordering of the authorship (pcw)
  • propose to model hasLanguage with dcterms:language [ a dcterms:LinguisticSystem ]
  • propose to model hasPrettyString with rdfs:label, dcterms:description
    • Isn't there also a prefLabel somewhere? (pcw)
  • discuss hasURL
    • what is the description of this property? Seems there are a bunch of existing properties that could be used (source, seeAlso, etc...) (pcw)
  • discuss hasUsageQueryContent
Facts about IW Meeting 2013-07-11RDF feed
Date11 July 2013  +
Personal tools