joao: Working on early warning
system, to monitor and alert emergency situations.
... Two standards in use: W3C Semantic Sensor Network Ont, and
Smart Appliances REFerence ontology (SAREF)
... I'm focusing on declarative model with rule bases.
... Once a decision is made, a broker sends out messages using
OASIS XML emergency message standard.
... Semantic IoT Early Warning Systems, an ontology
approach.
... Our research is to improve the semantic interop of IoT
Early Warning Systems.
... Theree problems identified: C1: Semantic integration of a
variety of data sources. Avoide loss of semantics
... C2: Processing in time- and safety-critical apps
... C3: Data analysis for effective responses. Avoid false
positives, enable high quality.
... Inter-iot is in its 3rd year, finishing in Oct or Nov. Goal
is to provide a framework to enable integration of different
IoT platforms.
... Want to use semantics in the outer layer, not only in the
app.
... At runtime you can represent the translation between
ontologies and perform them.
... We are collaborating in the app layer. Two main uses:
ehealth and logistics.
... Second part of our collaboration is semantic tanslation
between W3C Semantic Sensor Network Ont and SAREF.
... Requirements: IoT platforms should be able to coordinate
and share data. 5 use cases vehicle collision, detection,
hazardous health changes, temporal relations, wrong-way
driving, accidents with dangerous goods.
... Data sources include driver's ECG, HR, accelerometer;
Shimmer ECG 3; UniversAAL; SAREF, Hl7/aECG, FHIR;
... Data is translated into FHIR.
... Want to discuss with you SAREF for healthc.
... Want to exploit the role of semantic gateway, to enrich the
semantics.
... Example ECG measurement taken from FHIR site.
... Proposing an extension of SAREF for health.
... One option is, instead of creating new predicates in SAREF,
use predicates from FHIR/RDF.
... Importing only the predicates we need.
... Second option is to create a description (shown in UML) in
SAREF.
eric: Goal is to take advantage of FHIR/RDF to capture this?
joao: Yes.
eric: With streaming data?
joao: Yes.
eric: I mentioned FHIR W5 ont. Did you have use cases where that would payoff in short term?
joao: This would be the use case.
I'll show you the problem when using FHIR.
... When I open FIHR and scroll down the list of property I was
amazed. Quite difficult for an ontologist to instantiate an
example.
harold: We have another task, for
FHIR semantics, to take these properties and assign them to
what they represent in the real world.
... There have been some interesting experiments and starts,
such as SNOMED CT have mapped some of the content to
FHIR.
... The Clinical Info Modeling Initiative (CIMI) has worked on
that.
... And there are other possibilities being discussed. A group
is talking about how we ...
... We have a common way to represent healthcare data, and
enough semantics the most folks are getting it close. But the
next step is what you h are discussing.
eric: Another level indicator of
this in the FHIR process. FHIR staarted out as a data model in
XML and JSON.
... The harmonization that is now happening, where we say what
the semantics are, is the fusion of RDF back into the HL7
authoring process, which allows others to benefit from this
harmonization.
... The ideal is that the FHIR model itself migrates to be more
semantically awre.
harold: To say that two things
are equivalent, we first have to have names for them. That's
what we have nw in that bit list of properties that you are
showing.
... We're not as far along as what you might hpe, but you can
record the info in a a consistent and well understood ay in
RDF.
joao: Curious about this process.
eric: It is very early. W5 is a first step.
david: We don't yet have an identified group with a clearly defined mission to do it. But everyone on this call would like to do it!
eric: We need use cases.
... All the decisions that you make are based in the
terminologies.
... My thesis is that most use cases that will prioritieze
unification of terms will be wokflow use cases.
joao: By the end of this year I
would like to proose to ETSI (Euro Telecom Standards Institute)
to a SAREF for health.
... Would be very nice to have.
david: Connection between SAREF and FHIR?
joao: SAREF can be extended, and
has already been extended for energy, env and buildings.
... david: Taking parts of FHIR and defining them as being a
health extension for SAREF?
... Yes.
eric: As we unify, there will be machine readable reps of the mappings.
harold: Grahame indicated that
the current W5 is due for a radical overhaul.
... Charlie Mead is thinking of getting involved in it.
... Joao needs to connect up with the main FHIR community.
david: Recommend connecting with zulip
harold: W5 is currently used to
organize the home page.
... We generate OWL for W5 from a table that Grahame
maintains.
... One thing worth doing would be to go through RDF tools
about how to get from FHIR /JSON to RDF, for example.
... Would also be good to introduce Joao to ShEx.
... Joao post questions and we will start to answer them.
ADJOURNED
This is scribe.perl Revision: 1.152 of Date: 2017/02/06 11:04:15 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: s/Joao Moreira/Joao ("Joni") Moreira/ Present: David_Booth EricP Harold_Solbrig Joao_("Joni")_Moreira Rob_Hausam Gopi No ScribeNick specified. Guessing ScribeNick: dbooth Inferring Scribes: dbooth Found Date: 20 Feb 2018 People with action items: WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]