See also: IRC log
<Anita> Is the phone line up? I'm not hearing anything...
<bobP> phone is up!
<Lena> zakim is misbehaving, though...
<mscottm> Zakim is overloaded
<Lena> can't call :(
<Lena> disconnects itself before I have a chance to put the code in
<mscottm> I had to enter 4257# twice (as usual)
<mscottm> @Lena: maybe wait until traffic slows a bit
<mscottm> @Lena: I'll Skype you in if you keep having problems.
<Lena> i'm in!!!
<bobP> ericP and Michel starting the agendas.
<Lena> i think zakim saw eric and was afraid :D
<bobP> Rich Boyce, U Pittsburgh is here.
<bobP> Michel: First and foremost, let's thank the co-chairs
<bobP> ... have built lots of different communities
<bobP> ... New chairs, Michel, Charlie Mead,...
<ericP> scribenick: ericP
<michel> use case template: http://tinyurl.com/626fnxj
<bobP> ... Have developed templates for use cases;
<scribe> scribenick: bobP
take it over, ericP!
Charlie: Work at NCI, involved w
... considerable interest at NCI and HL7 to address problems of interoperability across delivery
... and HC and basic science
... agreed that use cases must be grounded
... HL7 came to realize that specs need to be limited to interoperability
<George> such as an RDFS or OWL version of CCD?
Charlie: HL7 has interest in this
... on the wire formatting; there is immediate interest
... CDISC has been focussed on clinical trial, but have also been building repositories
... perfect storm of opportunities between their requirements and semweb
... hope for active participation among HL7, CDISC, semweb
Lena(?): Even bigger gap between clinical and research
scribe: is there intention to bridge this too?
Charlie Mead: Yes, translational medicine.
<mscottm> Lena pointed out that there is a bigger gap between clinical research and wet lab biology research (compared to clinical trial and clinical research)
scribe: NIH perfect example of
data used in both cinical care and research domains
... incredibly complicated. Example is cohorts.
... sounds siimple, but it's not. Barriers are semantic primarily!
<Anita> Sorry what are 'cohorts'?
cohorts of patients in clinical trials
Lena: Data are proprietary, we cannot get our hands on the data
<mscottm> cohorts are the patients that have been aggregated for study according to inclusion and exclusion criteria
Charlie: Increasing pressure from users to break down the walls
<Anita> Thanks BobP - is that like groups with similar features?
Charlie: semantic problems are the deep problems
Anita: see Scott's comment - he the man :)
Charlie: 3B euro project to move data
<Anita> semantic = deep problem; openness = big problem - major effort in Europe to solve this
<Anita> We work on the deep problem ;-)
Charlie: starting to give way, exposing security and granularity issues
<Anita> Right thanks MScottM - got it!
Charlie: semweb can point the way. Consumers are demanding
ericP: i2b2 folks have addressed
half of this.
... when we make clearer what the semantics are, then we will be able to give IRBs more confidence
Lena: Should we start constructing sparql queries
Charlie: Project asking for hcls involvement
ericP: i2b2 has started down this
path of using existing deployed ehr systems and leveraging
semantics off of that
... infrastructure is there to exchange thru i2b2
... there is a whole space where semantics are not quite deployed
... but not too far away. Controlled vocabs are hierarchical
... can turn a sparl query into a large i2b2 query
Lena: But no guarantees, creating
another silo here??
... too much on our plates? Have to think how to do
ericP: We must work faster to be future-proof
Scott: U. Colombia has imported
some microarray data from kidney transplants
... success or failure of transplant
... also U. Pitt has RDF data w sparql endpoints
... also Sage, spoke to John Wilbanks who set up meeting in 2000
Scott: Sage working on pretty
much all the same goals
... talking w FDA, comparitor arm data
... maybe engage Sage w our tech and people
<boycer> clarification - the data is de-identified clinical reports from one month of hospital care at University of PIttsburgh Medical center. Intention is for NLP research. We created an index of semantic concepts present in the reports but researchers would still have to request the data through a data sharing agreement.
Charlie: Agree with Lena completely, want to answer questions
Scott: finishing...Sage also project IMI in Europe w Charlie and Michel
<boycer> further clarification on PItt data:The Pitt dataset is also now in TREC for information retrieval research: http://www.dbmi.pitt.edu/nlpfront
Scott: John Wilbanks,
consent-to-research permit(?) from patients
... notion of co-define privacy attributes for data and consent agreements
... so can reason over the legalese
... seems like a very promising areas, goes hand-in-hand w social approach
... helping to create conditions where clinical data become more available
... IMI project, would like to see IMI people also participate in hcls
<Anita> Are we discussing use cases in this call? I have a few I'd like to offer up
ericP: New charter
... developed use cases before the charter, some of them listed
... charter similar to last one, but calls out access-control policy
... and NIH, HL7, euro, etc
<boycer> Anita: I also sent one to the pub-sem-web list this morning but did not see it post---did you see it
ericP: If people have any good contacts w Euro or Asian versions of NIH - please help
Lena: Each Euro country will have its own NIH
<boycer> Anita: (referring to a draft use case)
Anita: No Euro FDA either
ericP: FDA in US would give us entre?
<Lena> @boycer, it's Lena who's speaking, not Anita :)
Scott: There are a number of Euro agencies
<mscottm> Regarding Lena's comment about EHRs being different languages in Europe: that's what make's RDF ideal - once you have the EHR model in RDF, you can switch between language labels.
<boycer> sorry, understood -- saw "Anita" question about use cases above
ericP: Certain amount of new work
in new charter, Scott has brag sheet
... also Pistoia and groups w pharma and vendors to pharma
... can take their use cases
... ties back to how to keep attention; impose more rigor in picking and reporting work
<Anita> Yes there are two actual females on this call ;-)!
Michel: Basic idea - to give structure to use cases
<mscottm> Let's hear it for Semantic females! 8)
<Anita> And neither of us are the shy, quiet type...
Michel: organizing by
functionality, efficiency, expense, how far reaching
... how to actually achieve this: develop new tech, open and free and unsupported, etc
<Anita> We should start a Grrrls4RDF group or something, Lena...
Michel: conceptually, lots of things are simple, but then tools; Must get a grip.
<mscottm> Excellent idea - would boost our membership if you keep association with HCLS.
Michel: What trying to achieve;
different to do research article than write notes
... partners. Outside of hcls, bring to table or bring things to the table.
... impact of the work. Varies from science to training in tech
... all these are metrics to evaluate projects
<George> template looks great - i'm on the queue
<mscottm> Anita is speaking (from point of view of Scientific Discourse task force)
<Anita> thanks for explaining and sorry for cutting in line! doh...
<Anita> FYI - our tentative use cases are up at http://www.w3.org/wiki/HCLSIG/SWANSIOC/Actions/RhetoricalStructure/meetings/20111010#Tentative_use_cases_as_per_06_Oct_2011
George: is hcls interested in use cases from affordable care act, etc?
<mscottm> no prob Anita
<George> great - thanks
<michel> michel: yes!
ericP: Writing down use cases helps us to get started. Yes,
<Zakim> Anita, you wanted to talk about use cases
Anita: We have a list of use cases that we are planning
<Anita> Thank you Michel. Is there a good format for saying what/who we need?
Michel: Objective is a change in
... we have rephrased groups to clubs
<Anita> Michel: change in the organisation - rephrase the interest groups but creating task forces around the use cases
Michel: but now there will be use-case-task-forces w fixed timelines
<boycer> good change - that allows cross interest group tasks
<Anita> People who have a joint interest will regroup themselves; use cases have 6 - 8 months/a year life time
Michel: in that sense we want to
scale back the scope
... must be manageable. Finite lifetime.
<Anita> scale back scope of subgroups; use cases define bite-size chunks
<Anita> both in time, particapation, and scope
ericP: Chunking happens inside use cases perhaps
Anita: Five sets of use cases up,
starting year and a half ago
... 1 id of components of medical text for key claims and evidence
... mark up so that there is a single hypothesis
... compare hyps based on evidence, etc
... Impact, NLP and markup, does it scale?
... high risk , high payoff
... 2) collaboration for biomed, lower risk
... 3) ...
... 3) is academic
ericP: How do you compare use case value?
Anita: Up to participants.
ericP: "I can work on five different things."
Michel: Understanding what the (meta?) values are, to judge among use cases
<George> I'll take a look -
ericP: These are styles of use cases. Would be nice to align specific one and prioritize
Anita: Start with the data.
<Anita> How to prioritise use cases? Lena: start with the data!
Lena! Start with the data.
<Anita> No Bob, that's lena again... I know, it's difficult :-)
<mscottm> I agree with the principle of starting with the data (in case you haven't heard me say that before ;) )
<Anita> Ok - last two: 4. Mining Treatment Outcomes: Develop a generic, anonymised, multi-EHS compliant format and an information architecture that allows access to symptom/treatment data as a Linked Data source that can be made freely available and used for drug efficacy/outcome (meta)studies * Leads: Anita de Waard, Joanne Luciano * Partners: RPI, Elsevier
Michel: Encourage the interest groups to (align) with this
<Anita> If anyone's interested in working on either of those, drop me a note... the last one is a project we are working on with the VU and Philips.
<Anita> echochocho - couldn't hear a word
<George> thanks all, bye
<boycer> all I heard ws geeks
<Anita> Ok thanks! me too - what were you saying EricP? Just repeat yourself... or make art with it...
ericP: Can continuing geeking
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found ScribeNick: ericP Found ScribeNick: bobP Inferring Scribes: ericP, bobP Scribes: ericP, bobP ScribeNicks: ericP, bobP WARNING: No "Present: ... " found! Possibly Present: Anita Charlie DaveHau George George_Thomas HCLS Lena Michel P0 Scott aaff bobP boycer egonw__ ericP https iker joined matthias_samwald mscottm rkiefer scribenick You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 06 Oct 2011 Guessing minutes URL: http://www.w3.org/2011/10/06-HCLS-minutes.html People with action items:[End of scribe.perl diagnostic output]