Actions

Ontolog Forum

Revision as of 02:57, 15 April 2016 by imported>KennethBaclawski
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Ontology Summit 2016 Synthesis - Thu 2016-04-14

Session Co-Chairs: Michael Gruninger and Donna Fritzsche

Abstract

In this session, we will discuss the dominant themes that have been emerging from the previous Track Sessions. These ideas will form the outline for the Summit Communique.

Agenda

Conference Call Details

  • Date: Thursday, 14-Apr-2016
  • Start Time: 9:30am PDT / 12:30pm EDT / 6:30pm CEST / 5:30pm BST / 1630 UTC
  • Expected Call Duration: ~2 hours
  • Dial-in:
    • Phone (US): +1 (425) 440-5100 ... (long distance cost may apply)
      • ... [ backup nbr: (315) 401-3279 ]
      • when prompted enter Conference ID: 843758#
    • Skype: join.conference (i.e. make a skype call to the contact with skypeID="join.conference") ... (generally free-of-charge, when connecting from your computer ... ref.)
      • when prompted enter Conference ID: 843758#
      • Unfamiliar with how to do this on Skype? ...
        • Add the contact "join.conference" to your skype contact list first. To participate in the teleconference, make a skype call to "join.conference", then open the dial pad (see platform-specific instructions below) and enter the Conference ID: 843758# when prompted.
      • Can't find Skype Dial pad? ...
        • for Windows Skype users: Can't find Skype Dial pad? ... it's under the "Call" dropdown menu as "Show Dial pad"
        • for Linux Skype users: please note that the dial-pad is only available on v4.1 (or later; or on the earlier Skype versions 2.x,) if the dialpad button is not shown in the call window you need to press the "d" hotkey to enable it. ... (ref.)
  • In-session chat-room url: http://webconf.soaphub.org/conf/room/summit_20160414
    • instructions: once you got access to the page, click on the "settings" button, and identify yourself (by modifying the Name field from "anonymous" to your real name, like "JaneDoe").
    • You can indicate that you want to ask a question verbally by clicking on the "hand" button, and wait for the moderator to call on you; or, type and send your question into the chat window at the bottom of the screen.
    • thanks to the soaphub.org folks, one can now use a jabber/xmpp client (e.g. gtalk) to join this chatroom. Just add the room as a buddy - (in our case here) summit_20160414@soaphub.org ... Handy for mobile devices!
  • Discussions and Q & A:
    • Nominally, when a presentation is in progress, the moderator will mute everyone, except for the speaker.
    • To un-mute, press "*7" ... To mute, press "*6" (please mute your phone, especially if you are in a noisy surrounding, or if you are introducing noise, echoes, etc. into the conference line.)
    • we will usually save all questions and discussions till after all presentations are through. You are encouraged to jot down questions onto the chat-area in the mean time (that way, they get documented; and you might even get some answers in the interim, through the chat.)
    • During the Q&A / discussion segment (when everyone is muted), If you want to speak or have questions or remarks to make, please raise your hand (virtually) by clicking on the "hand button" (lower right) on the chat session page. You may speak when acknowledged by the session moderator (again, press "*7" on your phone to un-mute). Test your voice and introduce yourself first before proceeding with your remarks, please. (Please remember to click on the "hand button" again (to lower your hand) and press "*6" on your phone to mute yourself after you are done speaking.)
  • RSVP to gruninger@mie.utoronto.ca with your affiliation appreciated, ... or simply just by adding yourself to the "Expected Attendee" list below (if you are a member of the community already.)
  • This session, like all other Ontolog events, is open to the public. Information relating to this session is shared on this wiki page.
  • Please note that this session may be recorded, and if so, the audio archive is expected to be made available as open content, along with the proceedings of the call to our community membership and the public at-large under our prevailing open IPR policy.

Attendees

Proceedings

[12:43] ToddSchneider: The SCOPE model has evolved over the first version with the recognition that it's applicable beyond electronic networks.

[12:45] ToddSchneider: Semantics is just more difficult.

[12:45] LeoObrst: One issue is not necessarily the OWL ontology format (which has its issues), but the ontological analysis that goes on prior to developing an ontology, in any language.

[12:46] ToddSchneider: Leo, yes!

[12:46] JackRing: Donna addresses a key barrier. Was heist semantic interoperability? Need an ontology of SI. Then how to detect that SI is needed and how to know that it is happening. Takes more than an glossary.

[12:46] MikeBennett: Ontology presentation does not have to be technical-looking, since real people understand real things well enough, and can be taught about FOL in terms of set theory.

[12:47] ToddSchneider: For those interested in the on going development of the SCOPE model you can join our weekly meetings.

[12:49] MikeBennett: On slide 3: maybe it's not that you do or don't have the "right" ontologies, but lack a clear meta-ontology so we know which ones can or can't interoperate with which other ones?

[12:49] Ram D. Sriram: @Michael: I believe the Healthcare track folks talked about several successful implementations!

[12:49] ToddSchneider: What constitutes a 'right ontology'?

[12:50] LeoObrst: @MichaelGruninger: ontologies are not enough for semantic interoperability. Also needed are multiple vocabularies that map to the same/similar ontologies. Also, one needs to understand the compact between the ontology and its use cases/applications, i.e., the formal pragmatics of use/intent.

[12:50] JackRing: Systems are devised by humans. Semantic Integration in systems depends on semantic integration among the humans. Where is the ontology of 'semantic' and 'integration'?

[12:51] ToddSchneider: As the SCOPE model makes clear there are many 'dimensions' to interoperability. Semantic interoperability is just one sub-space of these 'dimensions'.

[12:52] MikeBennett: Michael's Slide 4 is right on the money. Unfortunately a lot of people in some ontology efforts, like to develop their ontologies independently, and explicitly reject the use of upper ontologies, partitions etc, as being too complex for them. Instead they want to do the kind of thing that works for a stand-alone semantic application.

[12:52] LeoObrst: @MichaelGruninger: Yes, coordination is good, i.e., an architecture for a larger integration/interoperability framework, as long as the ontologies are not developed by large committees.

[12:52] Donna Fritzsche: agreed Leo - the analysis and communication of that analysis is key.

[12:53] ToddSchneider: Michael, the 'ramblings' are helpful.

[12:53] Donna Fritzsche: Also - regarding lack of consensus - see B Smith slides where he points out that multiple approaches to solving a problem is useful and a normal part of technology innovation.

[12:54] Mark Underwood: About Michael's slide 3 - seems similar to the state of affairs in Domain Specific Languages

[12:58] Donna Fritzsche: Leo - agree with your second point - do you mean "understand the contract between ontology and use cases." - fully agree if that is your point.

[12:58] JackRing: Must provide for detecting errors (logic, arithmetic, semantic, semiotic) in any semantic model.

[12:59] Mark Underwood: @Donna- U mentioned that u think more recently trained folks are more capable of . . . did u say better tagging? Use of metadata mgmt?

[13:00] LeoObrst: @Donna: yes, compact as an agreement between ontology (and its developers), use case/requirements developers, and eventual application developers.

[13:02] Donna Fritzsche: @Mark - I actually meant that 21st century kids will understand more about tagging/etc because of their pervasive use of technology for social reasons. (not a big point - just trying to relate it to models that older folks grew up with - math, science...)

[13:05] Mark Underwood: @Donna - OK. I don't share that optimism. Reference Sherry Turkle's book on Reclaiming Conversation. Tagging a la hash tags is sloppy, non-canonical. An old lessons these new dogs don't want to learn. <end of rant>

[13:05] Mark Underwood: *lesson

[13:05] Donna Fritzsche: it was really a minor point - just trying to be a little optimistic.

[13:06] LeoObrst: @Gary: Upper/foundational ontologies are also largely understood in detail only by philosophically oriented ontologists, i.e., they are domain experts of the foundations. Ordinary domain ontologists generally don't understand the distinctions, nor the need for those.

[13:08] Mark Underwood: @Leo - probably true about the GrayBeards, but in the API world there is an economy in reuse which could attempts not to reinvent some fundamentals, e.g., reasoning about time, etc.

[13:09] JackRing: Is the Periodic Table of the Elements an ontology? and Semantic Interoperability Agent?

[13:09] Mark Underwood: Donna: Understood, but in a way it might be somewhat more important as representative of a mindshare, or its lack

[13:10] ToddSchneider: Michael, is there an inherent conflict between ontologies that can be reused (as is) vs. ontologies that meet specific operational needs? That is, are ontologies that solve specific problems too constraining to be reused (because they need to represent very specific details, hence overly constrain interpretation not valid outside of the particular application for which they were developed)?

[13:10] Mark Underwood: *could influence attempts

[13:11] Donna Fritzsche: @Todd - that's a very good point. It also bring up the notion of bridging ontologies as opposed to reuse of full ontologies.

[13:13] MichaelGruninger: @Todd: I don't think that there is an inherent conflict between reusability and domain-specificity. There can be domain-specific ontologies that are reusable for a wide range of problems i.e. horizontal reuse vs the typical notion of reuse which is more vertical (e.g. more generic = more reusable).

[13:13] MikeBennett: @Todd very much so. Which means that a very good ontology build for some specific purpose, is not necessarily re-usable for integration just because it's good.

[13:13] Donna Fritzsche: @mark - Not sure exactly what you mean, but its significant that many of my non-IT adult friends now know what metadata is (for instance) and many under 20 know what it mean to "tag something". That being said - there is along road between that and understanding an implemented ontology.

[13:14] JackRing: Slide 6 confuses Fit vs. Function

[13:15] Mark Underwood: @Donna, glad to hear that :)

[13:16] Mark Underwood: @Donna They may think they know metadata from Snowden

[13:16] ToddSchneider: Michael, in terms of using ontologies to solve a systems interoperability or integration problem, this requires something more specific than a domain ontology.

[13:18] ToddSchneider: Leo, Ram, your slide #2 is great. Now you just need to add in the insurance companies:)

[13:18] ToddSchneider: Leo, Gary, the lack of understanding of upper ontologies and how to use them is evidence that a lot more 'education' is needed.

[13:20] Donna Fritzsche: @Mark - Well the Snowden effect is part of it, but I have heard a non IT person use the phrase Metadata Scheme on a Facebook post (much to my surprise.) I just recall 10 to 15 years ago when customers had no familiarity with the term metadata. Its definitely small steps though.

[13:21] RaviSharma: Leo references too many on postcoordination but some of them to community here FYI:Extending the coverage of phenotypes in SNOMED CT through post-coordinationhttps://www.healthit.gov/.../PMTF_SNOMED_Article_2015-09-10.pdf

[13:22] RaviSharma: also http://drmichaelstearns.com/post-coordination-pre-coordination-codified-concepts/

[13:22] JackRing: Probably this dialog would benefit from consideration of Derek Cabrera's DSRP semantic framework. c.f., https://www.crlab.us I have no business relationship with this

[13:24] Donna Fritzsche: thanks for the reference Jack

[13:26] RaviSharma: Ram / Leo - Yes many of background slides are good and illustrative and were not shown in earlier presentation.

[13:27] BobbinTeegarden: @Jack yes, I echo Donna, thanks for the reference.

[13:31] Donna Fritzsche: Oracle also owns Endeca

[13:32] MichaelGruninger: @Todd: what is an example of an ontology that is more specific than a domain ontology?

[13:34] ToddSchneider: Michael, how are you interpreting 'domain'? I interpret it as something that encompasses more than one possible solution space.

[13:34] Donna Fritzsche: Ended reference: http://www.oracle.com/us/solutions/business-analytics/business-intelligence/endeca/overview/index.html

[13:34] RaviSharma: Mark and Todd: Cloud as silo-breaker, is a bit worrysome from security and also federated access and leakages when you break the silo?

[13:35] RaviSharma: lot more scrutiny of integration is required or is it already built in?

[13:35] LeoObrst: @MichaelGruninger: A sub-domain ontology ;). Seriously, consider a role-specific projection (subset) of a given domain ontology, which could be modeled as a situation kind, but could also in a separate ontology that then gets imported (or imported by some bridging/integration ontology).

[13:36] ToddSchneider: Ravi, security and use of ontologies are usually not discussed together.

[13:36] Gary Berg-Cross: @Mike to your Q "what is an example of an ontology that is more specific than a domain ontology?" An application ontology that may be within a domain. or an ontology for some vocabulary that expresses terms in a sub-domain as recognized by practitioners.

[13:36] MichaelGruninger: @Todd: I am referring to examples such as slide 10 from Bob Young's presentation -- domain ontologies for part versions, part families, features, manufacturing methods, resources ,...

[13:36] LeoObrst: ^"could also be in a separate ontology"

[13:37] JackRing: The question regarding 'subdomain' illustrate the confusion between set-theoretic structures vs. semantic clustering.

[13:38] Donna Fritzsche: @ravi - point taken and somewhat agree at least based on user perceptions. On example of safe cloud storage at a massive scale: http://www.cleversafe.com/

[13:38] JackRing: The cloud serves to propagate ontology errors e**n faster than before.

[13:38] LeoObrst: @Jack: I don't think so: both could be set-theoretic.

[13:40] RaviSharma: Todd - can we summarize what (beyond data and RDBMS related) advantages we can derive from use of ontologies?

[13:41] Gary Berg-Cross: From Wikipedia A domain ontology (or domain-specific ontology) represents concepts which belong to part of the world. Particular meanings of terms applied to that domain are provided by domain ontology. For example, the word card has many different meanings. An ontology about the domain of poker would model the "playing card" meaning of the word, while an ontology about the domain of computer hardware would model the "punched card" and "video card" meanings.

[13:41] Mark Underwood: @Jack True - scaling up means scaling up risk. Related: resilience tends not to be a design goal

[13:43] Mark Underwood: @Ravi - there is work in OMG to use ontologies to represent security models, which the Cloud Security Alliance discusses, but the protection mechanisms in place do not incorporate ontologies as either code or data.

[13:44] Donna Fritzsche: "How to manage complexity" (quote from speaker)

[13:45] JackRing: We manage complexity by model-based design which in turn make explicit which terms and meanings are used.

[13:45] Mark Underwood: Todd - The managing dependency problem is a good one. Do we/ you think this is a good spot for ontology?

[13:45] RaviSharma: Donna - great scalability by riding over IBM cloud but what about security, I am not expert but would like both flexibility to cross cloud boundaries provided proper authentication and authorization are seamlessly provided.

[13:45] BobbinTeegarden: @Gary Implications are that context drives the 'interesting' part of an ontology (a sub-ontological set?) that's useful in a situation. This gets back to usage/process/context defining a sub-ontology that's useful to do something. It could be architected to be a cached subset useful for a process task (transient?). Useful for system architects, but seem to disturb the ontologists ;0).

[13:46] Mark Underwood: @Ravi - Have a look at Cloud Security Alliance for some comforting language

[13:46] Donna Fritzsche: @Jack - I agree that is a proper solution. Very difficult to do for certain domains - which I think was Todd's? point.

[13:47] ToddSchneider: Mark, almost any aid would help the current means to manage software dependencies.

[13:48] ToddSchneider: Gary, the quoted definition sounds good. But, I think there's still a gap from a domain ontology to an application/problem solution specific ontology.

[13:49] RaviSharma: Mike -Reporting / compliance and risk management - Are FIBO efforts in Sync with OMG and others risk ontology initiatives.

[13:49] Donna Fritzsche: @Bobbin - agreed.

[13:49] RaviSharma: Mark - I plan to look at CSA thanks.

[13:49] MichaelGruninger: It's interesting to compare the challenges in Tracks like GeoSciences and Engineering (where there are many ontologies) with Tracks like Finance (where people are using very few ontologies)

[13:50] Gary Berg-Cross: @Bobbin "Useful for system architects, but seem to disturb the ontologists ;0)." I think an ontology should serve its users who have a domain of expertise and define the scope of an ontology as well as its Universe of Discourse. We KEs help add quality to the semantic models.

[13:50] Donna Fritzsche: @Ravi - I believe they have a smart/encrypted storage across distributed across several nodes .. making it more secure - but you will have to read about it, my knowledge is out-of-date on Cleversafe (just recently bought by IBM).

[13:52] Mark Underwood: @Todd, sorry about the deck fork

[13:54] Mark Underwood: Which ISO did Mike just name-drop?

[13:54] JackRing: Does anyone know of an ontology of the kinds of '-nyms' that a multi-user ontology must contain?

[13:55] ToddSchneider: Jack, Yes, model based engineering is a step toward managing complexity by forcing the explicit representation of entities involved. But I've yet to see that they adequately constraint interpretation of the natural language terms used in the model. But their use is a step forward. The down side is that most current tools rely on UML and its profiles. UML is an impediment to having models that meet a goal of semantic explicitness.

[13:55] RaviSharma: Mike - In how many ways (recommended methods) can we apply Use of vocabularies, business glossaries and ontologies and tie Ontology to the first two?

[13:55] Gary Berg-Cross: Most extant domain ontologies aren't suitably grounded (sound), broad or deep enough to be reference ontologies.

[13:56] RaviSharma: Mark I looked at CSA and standards / certifications, will look more. appreciate.

[13:57] Gary Berg-Cross: Sorry, have to run so I will miss the discussion.

[13:57] Donna Fritzsche: Thanks for the great input Gary!

[13:58] ToddSchneider: I also need to go, unfortunately. Cheers.

[13:58] Donna Fritzsche: Thanks Todd!

[13:58] Mark Underwood: Thx Gary and Todd

[14:00] RaviSharma: todd and Gary thanks

[14:01] LeoObrst: I have to go too, folks. Thanks for the great points.

[14:01] Mark Underwood: Ciao Leo

[14:01] BobbinTeegarden: @Todd re UML impediment -- take a look at VOM Visual Ontology Modeler by Elisa Kendall, uses UML (ODM) for basis of good ontology; and CCM Cameo Conceptual Modeler, which is a more conceptual OWL modeling tool. Other UML tools are striving to produce more ontologically correct (UML-type) models.

[14:02] JackRing: Todd, Yes, the last decade in model-based system engineering has shown us what we should be doing. Those able to set current methods and tools aside and build the next generation will advance the practice considerably.

[14:03] Donna Fritzsche: @Bobbin - thanks for the modeling reference. Sounds interesting.

[14:05] Mark Underwood: Pretty skeptical about tools that don't work with Visual Studio, Eclipse, JetBrains. They tend to support analyst whiteboards, but remain apart from production code, hence have little value for maintenance

[14:06] JackRing: Leo, Pls define the 'set' that contains Roy Rogers, Dale Evans and Trigger. In today's usage can a set be based on what things do as well as on what they are?

[14:11] MikeBennett: @Mark hh:54: ISO 20022

[14:11] Mark Underwood: @Mike, thx

[14:12] RaviSharma: Mike - ETL at some synchronizing Epoch if asynchronous sources of info-data are used for ontology integration.

[14:13] MichaelGruninger: What are general principles and techniques for semantic integration that are common across all of the Tracks, and which aspects integration seem to arise in specific Tracks?

[14:14] RaviSharma: All- can we have concepts similar to situational awareness (e.g. maritime domain awareness which is also being used for NEXT GEN Airspace?)

[14:14] RaviSharma: This would allow concept and context mapping easy(??) any comments?

[14:14] JackRing: Michael, Primary gap is the lack of attention to the quality, e.g. Zero Defects, of an ontology.

[14:16] Mark Underwood: Michael - I see a theme in upper ontologies and reuse. Do these principles harken back to our earlier summit?

[14:17] Mark Underwood: @Ram - I gently resent the implicit ageism. There are plenty of young technophobes

[14:18] RaviSharma: This concept would track history and reconstruct context if there is discontinuity.

[14:18] JackRing: Michael, The second gap is the lack of specification of the dynamic and integrity limits of any given ontology (viewing an ontology as a system and applying these system principles).

[14:18] MikeBennett: @Mark hh:16 I certainly think so. The challenge is that for many very smart tech folks, because this calls upon a non-technical discipline, they seem to find it difficult (and then assume others will to). so there is a challenge in any multi-ontologist effort, in persuading people to follow those findings. IMHO.

[14:19] RaviSharma: May be we can call it semantic awareness?

[14:19] MikeBennett: +1

[14:19] BobbinTeegarden: @Mark regarding support for production code, the UML tools today can generate running code and do process simulation, other full cycle development. So if you create an ontology in the same model that models, and generates the running code (MBSE Model Based Systems Engineering), you end up with process that directly uses the ontology in the combined model.

[14:20] Ram D. Sriram: @mark: Ok I stand corrected

[14:21] Mark Underwood: Ram - But you are correct that there are cultural, sociological problems that are impeding progress

[14:22] JackRing: Mark, And you can end up with a process that has many unintended consequences. Be wary of any software which license contains the phrase 'not warranted for use'

[14:23] Ram D. Sriram: @Mark: I agree. We need to look at the acceptance of EHRs from cultural and sociological perspectives.

[14:23] Donna Fritzsche: HI Ravi, If you get a chance to right a few sentences about what you mean about "semantic awareness" - that would be helpful. Its an interesting phrase but I am not sure that I understand your intent.

[14:26] MikeBennett: The key thing for me was something that was documented in last week's session, that an ontology for reusability / integration needs to have a good foundational grounding, broad coverage and detailed axiomatization etc. (I borrowed this and put it in a slide, with attribution).

[14:26] Mark Underwood: The Upper Ontology Summit was 2006

[14:26] RaviSharma: will do that Donna - will write a para, it will also help temporal and spatial coherence at the least but may also address concatenation and telescoping of concepts.

[14:26] Donna Fritzsche: great, thanks.

[14:26] MikeBennett: In Finance there are relatively few ontologies, but given the breadth of these, the integration issues remain, but between the pieces.

[14:26] Mark Underwood: @Ravi it is a grt family of use cases, esp for sit awareness (DoD loves this one)

[14:28] JackRing: That sound you just heard was C.S. Pierce --- rolling over in his grave.

[14:28] Donna Fritzsche: Understand the domain culture, aids in attitudes towards acceptance/adoption. (speaker quote).

[14:28] Donna Fritzsche: Agreed with speaker.

[14:29] Donna Fritzsche: We did try to get a few speakers who explicitly address socio-technological issues - but they were too busy working to have time to speak!. I will post some references to their work when I get a chance.

[14:29] Mark Underwood: Pierce - whoa

[14:30] RaviSharma: Yes I worked on some of these at DOD and MITRE has lead the concept (MDA) and I worked on promoting it to Next Gen FAA and this kind of thinking has promoted auto-navigation by being aware of other airborne identifiable entities. (ATS) Transponders are required for all aircraft (med and large) regulation now.

[14:30] Donna Fritzsche: Ram say s- where do groups/domains fall on an ontology-use scale. Good idea

[14:30] MikeBennett: Ontological Maturity - great idea. I think this would apply not only to whether people are using ontologies, but how semantic is the work they are doing with those ontologies.

[14:30] Mark Underwood: Ram - suggests notion of Ontology Maturity Model

[14:31] Donna Fritzsche: better said Mark

[14:31] Donna Fritzsche: and Mike

[14:32] KenBaclawski: It would be useful to mention issues such as the early adopter problem. Domains that have committed to relatively weak semantics (such as markup languages) may now be reluctant to upgrade to richer semantics.

[14:32] RaviSharma: great session

[14:32] Donna Fritzsche: Thanks all!

[14:32] RaviSharma: ram any parting comments we are still on chat

[14:35] KenBaclawski: The idea of an Ontology Maturity Model would be useful, but the idea that a domain could gradually evolve toward a more mature use of ontologies may not be feasible in practice. Are the various maturity levels backward compatible?

Audio Recording

Session recording