Joint Ontolog-OOR Panel Discussion Session - Thu 30-July-2009    (1Y4T)

Title: BioPortal and related Ontology Repository Implementations and Development    (1ZOR)

Session Chair: Dr. NatashaNoy (Stanford) - [ opening slides ]    (1ZOS)

Panelists:    (1ZOT)

Archives    (204B)

Conference Call Details    (1ZJP)

Attendees    (1Z9Q)

Abstract:    (1ZA7)

Title: BioPortal and related Ontology Repository Implementations and Development    (204I)

This session will cover issues related to development efforts that are going into the BioPortal and the initial implementation of the OOR; particularly as it relates to software efforts as a contribution to the codebase of our first instance of the OOR that is built upon the NCBO BioPortal technology. - Chair: NatashaNoy - [ opening slides ]    (1ZA8)

Here are the abstracts of the briefings from our panelists that will seed our discussion:    (2080)

Resources:    (2082)

Agenda    (204X)

1. Opening by session Chair (NatashaNoy)    (204Y)

2. Briefings from Panelists -- CarlosRueda et al., StephenGranite, AlexGarcia, KatherineGoodier, KenBaclawski, MichaelGruninger & MikeDean    (204Z)

3. Q & A and Open Discussion (All) -- please refer to process above    (2050)

4. Summary and Next Steps (NatashaNoy)    (2051)

Proceedings    (2052)

Please refer to the archives above    (2053)

IM Chat Transcript captured during the session:    (2054)

 (The chat transcript below has been lightly edited to help improve on clarity of the conversation.)    (20NY)
 PeterYim: Welcome to: Joint Ontolog-OOR Panel Discussion Session - Thu 30-July-2009    (20NZ)
 Title: BioPortal and related Ontology Repository Implementations and Development    (20O0)
 Session Chair: Dr. NatashaNoy (Stanford)    (20O1)
 Panelists:    (20O2)
    * Dr. CarlosRueda, Mr. PaulAlexander & Mr. JohnGraybeal (MBARI) - "MMI Ontology Registry and Repository" 
    * Mr. MikeDean (BBN) - " OOR Progress, Federation, and Other Thoughts" 
    * Mr. StephenGranite (Johns Hopkins) - "Utilizing NCBO Tools to Develop & Use an ECG Ontology" 
    * Dr.  AlexGarcia (Bremen) - "The Ontology Repository for Assistive Technologies, ORATE" 
    * Professor KenBaclawski (Northeastern) - "OOR at Northeastern" 
    * Professor MichaelGruninger (Toronto) - "Common Logic Ontology Repository" (in absentia)    (20O3)
 PeterYim: session page is at:    (20O4)
 PeterYim: slides are at:    (20O5)
 anonymous morphed into Randy Kerber    (20O6)
 anonymous1 morphed into RexBrooks    (20O7)
 anonymous morphed into StuartTurner    (20O8)
 anonymous morphed into MichaelDorf    (20O9)
 anonymous1 morphed into MyCoyne    (20OA)
 anonymous morphed into KurtConrad    (20OB)
 anonymous morphed into ArturoSanchez    (20OC)
 anonymous morphed into StephenGranite    (20OD)
 anonymous1 morphed into AlexGarcia    (20OE)
 anonymous morphed into PavithraKenjige    (20OF)
 PaulAlexander: The BioPortal GUI deployment document I was referring to is located 
 here:    (20OG)
 RaviSharma: MikeDean, What is the OMV, ontology metadata view? based on what?    (20OH)
 AlexGarcia: OMV = ontology metadata vocabulary    (20OI)
 JohnGraybeal: Ontology Metadata Vocabulary, a vocabulary which provides a particular set of metadata 
 relevant to ontologies.    (20OJ)
 PeterYim: @ RaviSharma - see PeterHaase's OMV presentation 
 at:    (20OK)
 RaviSharma: John,Peter and Alex Thanks. Mike What is the end objective, at least like vocab. term, owl 
 or RDF links and integration towards executable ontologies?    (20OL)
 RaviSharma: already asked question on voiceline. thanks.    (20OM)
 AlexGarcia: not sure, the OMV just describes some basic features of an ontology. for instance was a 
 methodology followed? there are some basic very basic versioning descriptive features as well...    (20ON)
 RaviSharma: Alex- Can this be used for Concept Matching?    (20OO)
 AlexGarcia: yes it can    (20OP)
 RaviSharma: Stephen - Can you go beyond synthetic ECG (Novel) that have physical occurence in actual 
 patient databses that then can be matched to preexisting trends before severe events such as 
 fibrillations occur?    (20OQ)
 RaviSharma: Stephen - how will you find reliability of synthetic waveforms, by any occurences of such 
 forms in exisiting databases or during trace of events relating to cardi vascular abnormalities - 
 various grades    (20OR)
 StephenGranite: @ RaviSharma - that is part of the plan of the CVRG itself.  The ECGs visualized are 
 actual subject ECGs, derived from primary de-identified ECG data files.  We have referenced and placed 
 the Minnesota Code Classification System concepts in the ECG Ontology, so that annotations to the 
 waveforms by cardiologists can utilize a standardized annotation method.  These annotations are 
 persisted at the current time in a relational database for prototype purposes, but we have an object 
 model with future plans for XML database persistence.  Within the CVRG, we have several data services 
 that span relational and XML databases.  Information on those can be found in the CVRG wiki 
 (    (20OS)
 AlexGarcia: users are able to compare, visually, up to 3 ontologies    (20OT)
 JohnGraybeal: why is a SOAP interface important to the OOR registry?    (20OU)
 AlexGarcia: the idea is simple, via P4 users query several repositories at the same time. they are then 
 presented with some basic info for those ontologies part of the resulting set. users then select those 
 they would like to look in more detail and then they can visually inspect them.    (20OV)
 AlexGarcia: ok, thanks everybody but I will be running out of battery any time soon....    (20OW)
 PeterYim: Ref. Ken's plan to use the OOR for their students ... question for NCBO folks, any idea when 
 will we start running into scalability issues?    (20OX)
 JohnGraybeal: has there been any discussion of how to integrate all the student feedback for OOR's 
 registry with the activities of the BioPortal (and other) registry teams?    (20OY)
 CarlosRueda: @NatashaNoy & BioPortal team: are there/will there be any schema for facilititaing the 
 client-server communication using the REST services?  (something similar to SOAP, but for REST)    (20OZ)
 ArturoSanchez: Question for Ken: to what NSF program did you submit the DataNet proposal and if you can 
 elaborate on it and why did they find it to be not a good match ...    (20P0)
 ArturoSanchez: OOPS ... sorry ... offline?    (20P1)
 KenBaclawski: Arturo: Either send me email or attend the session next week.    (20P2)
 ArturoSanchez: @Ken: Thank you. I'll follow up with an email.    (20P3)
 RaviSharma: Ken - Why are we not connecting repository metadata to domain specific ontologies 
 semantically? at least through the Vocabularies and or Terms?    (20P4)
 anonymous morphed into TomRuss    (20P5)
 RaviSharma: Ken - I guess I would like to get clarity on "Annotates" legend in slide 4 - does that mean 
 a marker or collection?    (20P6)
 RexBrooks: For Natasha: Is there any plan to provide a SOAP-based interface that would allow the 
 BioPortal to be included in a structured SOA Service aggregation?    (20P7)
 PeterYim: Question for the panel (or the participants) ... Is anyone already doing, or have anyone 
 thought of doing, generic front ends to allow xml schemas and DTDs, maybe even database schemas to be 
 translated and accepted into the OOR? ... this, preumably, will expand the OOR's utility tremendously 
 (especially when a mapping service is available)    (20P8)
 JohnGraybeal: @ PeterYim -- A lot of projects are working on the kind of conversions you describe, some 
 with impressive success I think. Unfortunately don't have links in hand, but I think it is clearly going 
 this direction.    (20P9)
 JohnGraybeal: Unfortunately I will have to excuse myself in a few minutes.  Thanks all for the 
 discussion/presentations. (I will listen later for the answer to my question about "integrating feedback 
 and modifications back to BioPortal/other developments".)    (20PA)
 RaviSharma: Rex- do you mean orchestration or SOAF?    (20PB)
 RexBrooks: Actually I mean structured choreographies that would allow the future resources I expect to 
 be included in the BioPortal to be available in Emergencies where an ontology of symptoms plotted 
 against diagnoses and outcomes could be consulted in an emergency.    (20PC)
 RaviSharma: Rex- very interessting, will you be creating orchestration among services based on semantic 
 or ontology based rules?    (20PD)
 RexBrooks: For my purposes, I can use REST for one-off needs, but for pre-arranging aggregations of 
 services, I suspect SOAP may be easier to allow policy-based aggregations.    (20PE)
 RexBrooks: Not orchestrations, which have no flexibility because the services are controlled/hard-wired 
 into rigid workflows.    (20PF)
 MikeDean: I recall that WSDL can express both SOAP and REST.  Are there better standards now for a 
 formal specification of services that can support both?    (20PG)
 BobbinTeegarden: Any such thing as REST nesting?    (20PH)
 RexBrooks: But yes, definitely ontology-based rules. We are working on this in Emergency Data Exchange 
 Language Reference Information Model.    (20PI)
 RaviSharma: Rex - I am interested in understading how SOA relates to ontologies, not only as a tool but 
 as solution architecture?    (20PJ)
 RexBrooks: These would be somewhat flexible, heuristics based decision trees.    (20PK)
 RexBrooks: That's what we've been having weekly knock-down drag-outs over for more than 3 years in the 
 OASIS SOA Reference Architecture Subcommittee*. We're almost ready for the second 60-Day Public Review.
 [ *OASIS SOA Reference Architecture SC is part of the SOA Reference Model Technical Committee ]    (20PL)
 RexBrooks: We have to get the reference architecture in place in order to get to sound solution 
 architectures.    (20PM)
 RexBrooks: But we are aligning with OMG SoaML and The Open Group SOA Ontology.    (20PN)
 RaviSharma: Rex - I will go to  OASIS SOA and study it, as activie participant in Elissa's efforts on OMG ODM I am 
 still grappling on SOAML integration with OWL and UML.    (20PO)
 PaulAlexander: I think this raises good questions, will OOR be handling separate releases, bug 
 tracking, feature requests, etc?    (20PP)
 RexBrooks: Yup, we're all working on integrating XML, RDF, OWL and UML. Not easy.    (20PQ)
 RaviSharma: Rex- along with this the translation or inter conversion (interoperation) standards such as 
 XMI version upgrades become key? do you agree?    (20PR)
 RexBrooks: Yes, Ravi. We were talking about that this morning.    (20PS)
 PeterYim: Natasha suggested - (for OOR sandbox) report technical/support issues to the [oor-dev] listserv:  ... and (for BioPortal) to:    (20PT)
 PeterYim: eventually, when we have large number of users, we will have a [oor-users] listserv, I bet    (20PU)
 PeterYim: I will create this [oor-users] listserv before mid-Sep (that is when classes start for 
 Ken's student    (20PV)
 RaviSharma: Rex - please keep me on your email lists on interoperable standards and emerging SOA ones.    (20PW)
 PaulAlexander: I would really like to see something like Google Code or Gforge available for OOR    (20PX)
 PaulAlexander: Especially if OOR releases will be separate from BioPortal releases. We'll have our own 
 issues to deal with.    (20PY)
 RexBrooks: In the OASIS EDXL-RIM SC we're building the model behind the scenes and as the technical 
 foundation for the specifications we will be producing, combining the use of Protege and Enterprise 
 Architect.    (20PZ)
 RaviSharma: Appreciate and value your approach Rex.    (20Q0)
 RexBrooks: Will be in touch.    (20Q1)
 RexBrooks: Your opinions are valued.    (20Q2)
 RexBrooks: highly.    (20Q3)
 RaviSharma: Rex- with these efforts i would think that we will get best of promise of SOA and Model 
 Driven architectures both!    (20Q4)
 RexBrooks: That's one of my main goals.    (20Q5)
 RexBrooks: Extra cycles?    (20Q6)
 RexBrooks: Is that related to that thing called vacation?    (20Q7)
 RaviSharma: Rex- Great, very worthwhile goal. hope the participants also see it as important, of course 
 partly demonstrated in the work today, but Services were more of enablers on OOR and i am thinking that 
 Solutions use both SOA and Ontology as integrated capabilities to make each other happen in a meaningful 
 way.    (20Q8)
 RexBrooks: IN EDXL-RIM yes, in SOA-RM-RA not so much.    (20Q9)
 RexBrooks: Getting past class diagrams is pulling teeth.    (20QA)
 RaviSharma: Rex- I agree, with attraction of MOF how do we go beyond BPM type UML2 i.e. user friendly 
 constructs and ontological frameworks that integrate?    (20QB)
 RexBrooks: hmmn, good question. I don't have a ready answer. I'm still trying to get more formal UML 2 
 accepted in the standards process. Even though its not new, it tends to intimidate people.    (20QC)
 RaviSharma: Thanks everyone    (20QD)
 RexBrooks: Thanks indeed everyone!    (20QE)
 PaulAlexander: Thank you all, look forward to next time.    (20QF)
 RaviSharma: Rex- yes it does but conceptually i think it should be friendly in notation like BPMN2 and 
 then we can remove some of the inhibitions, I am also not up dated on all profiles in UML!    (20QG)
 PeterYim: inspired by Paul's remark just now ... we should start addressing the "Gatekeeping" issues 
 ... which we have ignored in the OOR-sandbox, so far ... but as we are near to deployment the OOR-
 production instance, that would be one thing that needs to be in place ... discussion today? .. or next 
 week?    (20QH)
 PeterYim: great session ... thank you Natasha, thanks to all panelists and to everyone for your 
 participation and contribution!    (20QI)
 PeterYim: session adjourned 12:30pm PDT    (20QJ)

Audio Recording of this Session    (2058)

For the record ...    (205H)

How To Join (while the session is in progress)    (205I)