HCLSIG/Meetings/2009-03-05 Conference Call
- Date of Call: Thursday March 5, 2009
- Time of Call: 11:00am Eastern Daylight Time (EDT), 16:00 Greenwich Mean Time (GMT), 17:00 Central European Time (CET)
- Dial-In #: +1.617.761.6200 (Cambridge, MA)
- Dial-In #: +33.4.89.06.34.99 (Nice, France)
- Dial-In #: +44.117.370.6152 (Bristol, UK)
- Participant Access Code: 4257 ("HCLS").
- IRC Channel: irc.w3.org port 6665 channel #HCLS (see W3C IRC page for details, or see Web IRC)
- Duration: ~1h
- Convener: Scott, Susie
- News Updates by BioRDF, COI, LODD, Pharma Ontology, Scientific Discourse, Terminology - All
- C-SHALS trip report - Susie
- Intro SKOS, Summary of SKOS developments - Alistair Miles (Oxford)
- Use cases for representing data in SKOS vs OWL - Eric
Attendees: Scott, EricP, helena, johnM, MaryKennedy, CoryH, curoli, Kingsley_Idehen, Matthias_samwald, Bosse Andersson, jun, KevinDoyle, Ratnesh, SimonJupp, seanb, Kei_Cheung, EricN, alistair Miles, mroos
..following up on the query federation work, we are considering the Special Issue of BMC Bioinformatics, contacted editors about deadline extension to give us a fighting chance.
..Need to think about the F2F meeting in the BioRDF group.
Eric (COI): showed demo at C-SHALS about patient eligibility making use of pushing SPARQL queries through to mySQL databases
..Helen, Holger, and Eric are still working on the demo.
Bosse (LODD): got paper back, working on incorporating the reviewer's comments, need to think about F2F.
John (PharmaOnto): garbled (sorry!).
NIH Challenge Grants announcement: http://grants.nih.gov/grants/funding/challenge_award/Omnibus.pdf (esp. pp 128ff)
John (Terminology): had a good presentation at C-SHALS, AIDA is being extended with new features, MeSH SKOS has been added to server.
johnM (AOB): Check this out, relevant to HCLS IG: http://grants.nih.gov/grants/funding/challenge_award/Omnibus.pdf
..it's a large collection of NIH grant descriptions
-> http://grants.nih.gov/grants/funding/challenge_award/Omnibus.pdf call for NIH Challnge Grants proposals
-> http://www.w3.org/2009/Talks/0205-skos-egp/ skos use case slides
Intro (HelenaDeus): Grad student at UTexas interested in SW use cases
I have joined these meetings a few times, but it has been a few months since I was last on a call
KevinDoyle: platform for SW systems for lifcesci market
... on top of 11g
Intro (Sean Bechhofer, UManchester, involved in SW for many years - [scott paraphrases]
aliman: Alistair Miles
... work with flyweb at Manchester
... worked with SWBP, developing skos
I think flyweb is from Oxford
topic: skos intro [ aliman ]
Intro (Simon Jupp): Comes to us from UManchester, works for Robert Stevens, can provide information about SKOS tools (worked on SKOS editor for Protege)
aliman: used to describe existing classification schemes which are not full-fledged ontologies
... useful when creating a new ont when you just want to provide navigation framework
... last call WG Aug or Oct last year
... last call WG 29 Aug 2008
... waiting i18n to submit comments
... hope for CR very shortly -- following a call for implementations
-> http://www.w3.org/2006/07/SWD/SKOS/reference/20081001/ (SKOS editor's draft)
... excited to hear about our use of skos
johnM: most folks here are using 2004 skos
-> http://www.w3.org/TR/2005/WD-swbp-skos-core-guide-20051102/ 2005 SKOS Core Guide
aliman: 2005 SKOS Core Guide is about to go to candidate rec
... basic overview hasn't changed
... new skos abstract synopsis breaks down the skos modules:
... .. concepts
... .. labeling with plain literals in natural language
... .. notations -- lexical code
... .. 
... .. mapping between concepts
... .. grouping into concept schemes
... mapping and notations modules are new
aliman: SKOS is broken into many modules: e.g. language, concept schemes, mapping, document properties, Semantic relations, etc.
... appendix skosxl for describing more about lexical entities
... RDF literals can't be subjects in RDF
... you either need to do some tricks, or identify them in another way
... skoxsl can say that "x" is a transliteration of "y"
... defn for skos:broader was transitive in 2005
... no longer transitive, but has a transitive super property
... eases getting at direct hierarchale relations
-> http://www.w3.org/2009/Talks/0205-skos-egp/ skos v.s. owl
topic: skos vs. owl
"Weakened Ontologies vs. Qulaified Ontolgies"
Evolving use-case for SKOS: as a central thesauri repository that can be used by text-mining applications to produce normalized (concept uri) output
... within one pharma company, this is producing usable terminologies many times more efficiently than if OWL was being used.
ericP (slide2): some rdfs reasoning that you can do, asks Alistair if you can do that with SKOS
is it safe to say: SKOS is to Phenotype what OWL is to Gentotype?
Alistair: answers yes, with skos:broaderTransitive
..can connect Fido with Animal via transitive relations
so SKOS is for ABox oriented categorization/classification (even folksonomy structure) and OWL for the TBox
SKOS best for descriptive forms where knowledge is limited, e.g., phenotypes and symptoms
FYI from http://www.w3.org/TR/2008/WD-skos-reference-20080829/#L1045 -- """
To make the "knowledge" embedded in a thesaurus or classification scheme explicit in any formal sense requires that the thesaurus or classification scheme be re-engineered as a formal ontology. In other words, some person has to do the work of transforming the structure and intellectual content of a thesaurus or classification scheme into a set of formal axioms and facts. This work of...
...transformation is both intellectually demanding and time consuming, and therefore costly. Much can be gained from using thesauri etc. "as-is", as informal, convenient structures for navigation within a subject domain. Using them "as-is" does not require any re-engineering, and is therefore much less costly.
note release of the latest LOD cloud bubble diagram: http://www4.wiwiss.fu-berlin.de/bizer/pub/lod-datasets_2009-03-05.html (clickable version) and (colored graphic)
SKOS -> OWL does appear to be a practical evolutionary approach within orgs
Alistair: two approaches: start with SKOS and eventually create an OWL copy, updating the mirrors
..2) is mix the models, start with SKOS, add rdfs relation between SKOS concepts, e.g. rdfs:subClass
Sean (?): I would like to see use cases for these approaches.
ericP: OWL punning isn't a problem in SKOS (slide6)
When it is superficial (phenotypical) errors aren't costly
or aren't as costly as when they are genotypical (OWL)
Sean: How do I know when the machine is telling me something wrong?
ericP: Would the machine have been able to make any decisions using SKOS anyway?
John: OWL will always be correct.
KevinDoyle has left #HCLS
anobody still there? eric mentioned there were more people interested in managing user permissions in RDF.