From OntologPSMW

Jump to: navigation, search
[ ]

Contents

Ontology Summit 2011: Panel Session-6 - "Integrating the Ontology Application Framework, Use Cases, Value and Metrics" - Thu 2011_03_03     (1)

Summit Theme: OntologySummit2011: Making the Case for Ontology     (1A)

Session Title: Integrating the Ontology Application Framework, Use Cases, Value and Metrics     (1B)

Session Chair: Dr. SteveRay (CMU)     (1C)

Panelists:     (1D)

  • Professor MichaelGruninger (U of Toronto) - "The Ontology Application Framework"     (1E)
  • Mr. RexBrooks (Starbourne) - "Value Metrics"     (1F)
  • Mr. MikeBennett (Hypercube) - "Use Cases - how they fit into the framework"     (1G)

Abstract     (1I)

OntologySummit2011 Theme: "Making the Case for Ontology"     (1I1)

  • Session Title: Integrating the Ontology Application Framework, Use Cases, Value and Metrics     (1I2)

This is our 6th Ontology Summit, a joint initiative by NIST, Ontolog, NCOR, NCBO, IAOA & NCO_NITRD. The theme adopted for this Ontology Summit is: "Making the Case for Ontology."     (1I3)

This year's Ontology Summit seeks to address the need to provide concrete evidence of successful deployment of ontologies by examining several application domains for such examples, and in better articulating where different "strengths" of ontological representation are best applied. To support that, the summit also aims to classify the categories of applications where ontology has been, and could be, successfully applied; to identify distinct types of metrics that might be used in evaluating the return on investment in an ontology application (cost, capability, performance, etc.); to lay out some strategies for articulating a case for ontological applications; and to identify remaining challenges and roadblocks to a wider deployment of such applications that represent promising application areas and research challenges for the future. The findings of the summit will be documented in the form of a communiqu�� intended for public consumption.     (1I4)

This session begins the process of combining the excellent work done within several of the tracks. Specifically, we will hear from Track 1 - Ontology Application Framework - where Michael Grüninger will describe how they have categorized various classes of applications that use, or could use, ontology. Then, Rex Brooks (Track 3 - Value Metrics and Value Models) will show how this fits into a larger context where the applications and measures of benefits form a context that includes the different kinds of stakeholders and how different benefit (or value) metrics can be of use to different audiences and with different kinds of applications. Finally, Mike Bennett (Track 2 - Documented case studies) will show how some of the use cases we have seen fit into these frameworks.     (1I5)

The second half of the session will be open for discussion and questions.     (1I6)

See developing details on this Summit series of events at: OntologySummit2011 (home page for this summit)     (1I7)

Agenda     (1J)

Ontology Summit 2011 - Panel Session-6     (1J1)

  • Session Format: this is a virtual session conducted over an augmented conference call     (1J2)

Proceedings     (1K)

Please refer to the above     (1K1)

IM Chat Transcript captured during the session    (1K2)

see raw transcript here.     (1K2A)

(for better clarity, the version below is a re-organized and lightly edited chat-transcript.)     (1K2B)

Participants are welcome to make light edits to their own contributions as they see fit.     (1K2C)

-- begin in-session chat-transcript --     (1K2D)

[09:27] Peter P. Yim: Peter P. Yim: Welcome to the ...     (1K2E)

Ontology Summit 2011: Panel Session-6 - Thu 2011_03_03     (1K2F)

Summit Theme: Ontology Summit 2011: Making the Case for Ontology     (1K2G)

Session Title: Integrating the Ontology Application Framework, Use Cases, Value and Metrics     (1K2H)

Session Co-chairs: Dr. Steve Ray (CMU)     (1K2I)

Panelists:     (1K2J)

Please refer to dial-in information, agenda, and other details on the session page     (1K2N)

anonymous morphed into Matt Hettinger     (1K2P)

anonymous morphed into Marcy Harris     (1K2Q)

Steve Ray: Cheat sheet: *2 to mute, *3 to unmute     (1K2R)

anonymous morphed into Randy Coleman     (1K2S)

Mike Bennett: Tried the magic thing     (1K2V)

Mike Bennett: I'll diall out and in again     (1K2W)

anonymous morphed into Naicong Li     (1K2X)

anonymous morphed into JulitaBermejoAlonso     (1K2Y)

Steve Ray: Sorry about that, Rex. Your voice is definitely clearer with the handset.     (1K2Z)

Nicola Guarino: Maybe we should specify "where" the functionality provided by ontologies is applied.     (1K2AA)

E.g, specification is applied to conceptualizations; classification can be applied to concepts or     (1K2AB)

indivduals...     (1K2AC)

Steve Ray: One of the "-ilities" that fell off is "capability"     (1K2AD)

Steve Ray: @MichaelGruninger: Interesting that you didn't choose "specification" in the Integration     (1K2AE)

category, as well as mapping.     (1K2AF)

Rex Brooks: @MichaelGruninger: I think of Decision Support as marshalling information more than     (1K2AG)

automated inferencing. However, if the inferencing marshall information, then I'm good with it.     (1K2AH)

Nicola Guarino: Information integration can also occur at development time, in the sense that     (1K2AI)

ontologies are used to manually integrate different conceptual models. In these cases, ontologies     (1K2AJ)

can also help to recognize the impossibility of integration...     (1K2AK)

Michael Grüninger: @Nicola: "Where an ontology is used" might indeed be another dimension in the     (1K2AL)

framework     (1K2AM)

Michael Grüninger: @Nicola: Although information integration can occur at design time, I put this     (1K2AN)

into the Ontology Augmentation system category because it is being used to integrate conceptual     (1K2AO)

models before the system itself is designed. Unless you mean that conceptual models are integrated,     (1K2AP)

in which case I wonder whether this is also being used at run-time     (1K2AQ)

Steve Ray: @MichaelGruninger: It would be great if you could add the answers you gave on the phone to     (1K2AR)

the wiki page, for each of the classes of applications, i.e. which dimension, who is the user, etc.     (1K2AS)

Randy Coleman: @MichaelGruninger: Have you had thoughts about multi-dimension/multi-classification     (1K2AT)

ontologies?     (1K2AU)

Michael Uschold: Let me walk people through this -     (1K2AV)

cationFramework.pdf     (1K2AZ)

Rex Brooks: @MichaelUschold: One thing I didn't think about was results v. value metrics. sometimes     (1K2AAA)

the results are unfavorable and we need to capture that and learn from it., too.     (1K2AAB)

Fabian Neuhaus: @ Michael Grüninger: are "problem addressed" and "benefit" different dimensions?     (1K2AAC)

Wouldn't being able to address a problem be a benefit?     (1K2AAD)

Michael Grüninger: @Fabian: Earlier discussions have indicated a need to distinguish the benefit an     (1K2AAE)

ontology can deliver from the original problem that motivated the use of the ontology in the first     (1K2AAF)

Matt Hettinger: @MichaelGruninger if a set of ontologies are used at design time for a system other     (1K2AAH)

than the ontology system that is being used (e.g. some line-of-business system) then the use of that     (1K2AAI)

set of ontologies, that ontology system, can be (should be) considered a run-time use.     (1K2AAJ)

Steve Ray: @MichaelUschold: Probably the best way to decide what perspective to take with your     (1K2AAK)

diagrams is to think about the original purpose of the Summit, which is to help people trying to     (1K2AAL)

make the case. If a given aspect doesn't speak to that, then you won't need to include it.     (1K2AAM)

Steve Ray: @MichaelUschold: We lost your audio. Can you dial back in?     (1K2AAN)

Michael Uschold: how long ago did you miss me?     (1K2AAO)

Steve Ray: @MichaelUschold: It was only recent. You were doing the benefits.     (1K2AAP)

Fabian Neuhaus: @MichaelUschold: I think you were about to finish your presentation (when you got cut     (1K2AAQ)

Peter P. Yim: @MichaelUschold - you should be back on now ... try your voice     (1K2AAS)

Pavithra Kenjige: Generally cost benefit is analyzed during Business Case development to get     (1K2AAT)

funding.. these are good points..     (1K2AAU)

anonymous morphed into Bart Gajderowicz     (1K2AAV)

Nicola Guarino: what is it Ajax?     (1K2AAW)

Jim Rhyne: Ajax is a browser programming technique that ships data to the browser in XML and uses     (1K2AAY)

Javascript to perform local interactions with the user, avoiding the need for roundtrips with the     (1K2AAZ)

http server.     (1K2AAAA)

Bart Gajderowicz: AJAX was developed by a UI designer, so its benefits are best demonstrated by     (1K2AAAB)

example. Google maps is great for that. Ontologies may not be as easily demonstrable. Actually, that     (1K2AAAC)

designer specifically coined the term AJAX because explaining the technology was not "selling" the     (1K2AAAD)

Jim Rhyne: @MichaelUschold - curious about the diagramming technique shown in e6owl Legend PDF. Can     (1K2AAAF)

you point me to additional documentation. I work on the OWL profiles for UML.     (1K2AAAG)

DJimRhyne: @MichaelUschold - never mind, found it on the Semantic Arts website.     (1K2AAAH)

JustinCote: I think this is why we like PowerPoints     (1K2AAAI)

Steve Ray: @MikeBennett: It looks like many of the case study providers gave benefits, but we're     (1K2AAAJ)

having difficulty getting all the way to actual metrics.     (1K2AAAK)

Mike Bennett: @SteveRay I agree, that is a challenge. I hope we can get some of the Case Study     (1K2AAAL)

presenters to come back and put metrics in, in line with what's now in the Metrics track.     (1K2AAAM)

Rex Brooks: @Steve: Sorry for jumping the gun on ya.     (1K2AAAN)

Steve Ray: @Rex: No problem at all.     (1K2AAAO)

Rex Brooks: @MikeBennett: Excellent work. Extracting these is no picnic.     (1K2AAAP)

Nicola Guarino: my audio got disconnected. Calling again.     (1K2AAAQ)

Michael Grüninger: @MikeBennett: I think that the Purpose of the ontology that I presented is     (1K2AAAR)

synonymous with your term "Challenges"     (1K2AAAS)

Michael Uschold: I also agree, ontology features describe the ontology, as @MichaelGruninger is     (1K2AAAU)

Steve Ray: @MikeBennett: Do you observe good coverage over all the dimensions and aspects that     (1K2AAAW)

Michael Grüninger laid out, or, just as interestingly, do you see clustering of use cases around just     (1K2AAAX)

Peter P. Yim: @JohnSowa - can you capture your point to this chat (so it will go into the transcript)     (1K2AAAZ)

Michael Uschold: @JohnSowa: I build the ontology mainly to help clarify my own thinking, as well as     (1K2AAAAA)

to have more precise artifact to communicate to others. I was going round and round thinking about     (1K2AAAAB)

many distinctions. It was easier and faster for me to draw a picture (which translated into OWL)     (1K2AAAAC)

than write yet more text in say Word or html. I agree that many will not care about this mode of     (1K2AAAAD)

presenting a framework.     (1K2AAAAE)

Jim Rhyne: @JohnSowa - is there no middle ground between inspiration and CYA?     (1K2AAAAF)

Bruce Bray: The rare disruptive technologies / "killer apps" don't need boring metrics, but the usual     (1K2AAAAG)

incremental improvements often do in order to get funded.     (1K2AAAAH)

Michael Grüninger: @JohnSowa: the issue is that we need a common way of specifying the use cases --     (1K2AAAAI)

what are the relevant pieces of information that need to be included? If people have different     (1K2AAAAJ)

expectations about what use cases are, then we have not made progress.     (1K2AAAAK)

Michael Uschold: I very much agree with Fabian's point, that we want systematic way to think about     (1K2AAAAL)

and communicate the use cases.     (1K2AAAAM)

Jim Rhyne: @JohnSowa - Steve Jobs is a brilliant risk taker who is supported by some of the brightest     (1K2AAAAN)

market, usability and technology analysts around. He does not invent iPADs in a vacuum.     (1K2AAAAO)

Michael Grüninger: We also need to battle Buzzword Bingo -- if someone says "I used an ontology for     (1K2AAAAP)

semantic integration" (or any other buzzword), we need everyone to agree on exactly what that means.     (1K2AAAAQ)

Otherwise, people will be expecting something specific and then be disillusioned when they don't see     (1K2AAAAR)

Michael Grüninger: @JohnSowa: Of course, if we had the one case (killer app) then we would use it.     (1K2AAAAT)

The problem is that people are using ontologies for many different reasons. Everyone has a different     (1K2AAAAU)

killer app that they are looking for.     (1K2AAAAV)

Matt Hettinger: @RexBrooks (and any other who may have an answer) With respect to measures / metrics.     (1K2AAAAW)

A valid measure, at least in the relational approach to measurement, is that it is required that     (1K2AAAAX)

there is a empirical model, a numerical model, and a mapping between the two. If these do not exist     (1K2AAAAY)

then what is called a "measure", strictly speaking, is not a measure. Once consequence is the level     (1K2AAAAZ)

of trust in the numbers produced. For the measures discussed, are there empirical models, numerical     (1K2AAAAAA)

models and mappings. (I'm not familiar with other approaches to measurement)     (1K2AAAAAB)

Rex Brooks: @MattHettinger: I agree, and yes, for measurements such as length of time for     (1K2AAAAAC)

implementation, response time in the logs for performance, uptime v. downtime, etc, there is a     (1K2AAAAAD)

correct and trustable relationship. However for qualitative measurements like Customer Satisfaction,     (1K2AAAAAE)

which I expect many of us would not believe if it was just thrown at us to take at face value, it is     (1K2AAAAAF)

more difficult. I'm not saying such qualitative measurements should be accepted. Quite the contrary.     (1K2AAAAAG)

I think they need to be dissected and analyzed before any trust is achieved.     (1K2AAAAAH)

Rex Brooks: Oh yeah, they have to be valid after the analysis, or else require a boatload of     (1K2AAAAAI)

corroborating evidence.     (1K2AAAAAJ)

Steve Ray: Risk is another metric we might have overlooked - certainly important to many decision     (1K2AAAAAK)

Mike Bennett: @Michael there is a real risk of this in some of the ISO WGs (which shall remain     (1K2AAAAAM)

nameless), where some technical folks are starting to describe semantics / ontology / OWL as some     (1K2AAAAAN)

kind of magic paint.     (1K2AAAAAO)

Mike Bennett: @SteveRay Definitely. At the business level, every business case comes down to one or     (1K2AAAAAP)

other of cost and risk. What else is there?     (1K2AAAAAQ)

Michael Uschold: This conversation is more about strategy than it is about integrating OAF, Value     (1K2AAAAAR)

Metrics and Cases.     (1K2AAAAAS)

Rex Brooks: @MichaelUschold: true, maybe you could put your hand and address "measuresBenefit?"     (1K2AAAAAT)

Mike Bennett: Interoperability (business problem): Conceptual model (solution)?     (1K2AAAAAU)

Matt Hettinger: The killer app, from my perspective, is (inter-)enterprise architecture, as in both     (1K2AAAAAV)

my research and EA practice, all of the use cases, to some degree, can be entailed by EA     (1K2AAAAAW)

Pavithra Kenjige: Many people did not like the abbreviation "OAF" .. can we do something about that?     (1K2AAAAAX)

John F. Sowa: I agree that interoperability is essential.     (1K2AAAAAY)

Bart Gajderowicz: I 2nd the need for explaining the benefits of "semantics". Often people reject     (1K2AAAAAZ)

semantics based on individual implementations like RDF. Before people realize the benefits of     (1K2AAAAAAA)

ontologies, I often hear that the same problem can be solved using existing and less complex     (1K2AAAAAAB)

technologies. I hear that even Siri "can be done" using data mining techniques.     (1K2AAAAAAC)

Ken Baclawski: @BartGajderowicz: Are ontologies the "more complex" technology (compared with data     (1K2AAAAAAD)

mining, for example)? If that is true, perhaps we need to address the way ontologies are perceived.     (1K2AAAAAAE)

Data mining is hardly a simple technology, and ad hoc solutions to interoperability are very     (1K2AAAAAAF)

Mike Bennett: @KenBaclawski one issue on perception: could also say that all apps have an ontology     (1K2AAAAAAH)

anyway, we are about managing ontologies, having an explicit ontology in some formal notation etc.     (1K2AAAAAAI)

i.e. simplifying the (existing) problem of ontologies     (1K2AAAAAAJ)

Michael Uschold: @KenBaclawski agree that ontology technology can reduce complexity in systems. Even     (1K2AAAAAAK)

if ontology technology is complex compared to other approaches (when viewed on their own), sometimes     (1K2AAAAAAL)

it takes a complex technology to simplify a greater whole.     (1K2AAAAAAM)

Ken Baclawski: @MichaelUschold: That is a good approach to dealing with the perception problem, but     (1K2AAAAAAN)

we need to try addressing it by a number of routes.     (1K2AAAAAAO)

Mike Bennett: Also conceptual model as an industry standard, which perhaps we haven't explicitly     (1K2AAAAAAP)

Pavithra Kenjige: Interoperability -? We need businesses that used Ontology to do the case study     (1K2AAAAAAR)

regarding interoperability//     (1K2AAAAAAS)

Nicola Guarino: A general question I have is: how the ontology quality affects the benefits achieved     (1K2AAAAAAU)

for each of these use cases?     (1K2AAAAAAV)

Rex Brooks: @Nicola: Yes, exactly. Ideas?     (1K2AAAAAAW)

Nicola Guarino: @RexBrooks: for example, in some cases a lightweight ontology is enough, in other     (1K2AAAAAAX)

cases it doesn't work at all...     (1K2AAAAAAY)

Michael Uschold: @SteveRay: IHMO the real reason there are few case studies with metrics is because     (1K2AAAAAAZ)

it is hard to have enough of the variables controlled to get any meaningful measures. Software     (1K2AAAAAAAA)

projects do not lend themselves to this.     (1K2AAAAAAAB)

Leo Obrst: In addition to benefits, you need to have costs, but also relative costs, which can be     (1K2AAAAAAAC)

measured with respect to other metrics. E.g., degree of precision. Technology A vs. Technology B vs.     (1K2AAAAAAAD)

ontologies may all eventually provide the same level of precision (of results, of transactions,     (1K2AAAAAAAE)

etc.), but require much different costs to do so.     (1K2AAAAAAAF)

Rex Brooks: @LeoObrst: Good points.     (1K2AAAAAAAG)

Mike Bennett: @MichaelUschold - that is a very good point. Some industries are not native to systems     (1K2AAAAAAAH)

development and won't have the metrics of what they did before, that did not work so well, by     (1K2AAAAAAAI)

Nicola Guarino: I was just discussing with Rex about the need to evaluate the various case studies     (1K2AAAAAAAK)

with respect to the quality and depth of the ontology they used     (1K2AAAAAAAL)

Leo Obrst: We've found that people who have not gone through the XML "revolution" do not understand     (1K2AAAAAAAM)

the value of semantics. When they adopt XML and find out it doesn't provide what they need, they     (1K2AAAAAAAN)

begin to understand the difference between structural/syntactic approaches and semantic approaches.     (1K2AAAAAAAO)

Mike Bennett: @LeoObrst - that exactly reflects the financial services industry experience     (1K2AAAAAAAP)

Bobbin Teegarden: one of the major breakthroughs using ontologies/semantic web is the ability to     (1K2AAAAAAAQ)

visualize whole systems and interactions, getting a 'holonic' view of systems of systems -- looking     (1K2AAAAAAAR)

for this at the granularity of 'application' may be missing the point     (1K2AAAAAAAS)

Matt Hettinger: @BobbinTeegarden - yes!     (1K2AAAAAAAT)

Peter P. Yim: - session ended: 11:10 --     (1K2AAAAAAAU)

-- end of in-session chat-transcript --     (1K2AAAAAAAV)

  • Further Question & Remarks - please post them to the [ ontology-summit ] listserv     (1K2AAAAAAAW)
    • all subscribers to the previous summit discussion, and all who responded to today's call will automatically be subscribed to the [ ontology-summit ] listserv     (1K2AAAAAAAW1)
    • if you are already subscribed, post to <ontology-summit [at] ontolog.cim3.net>     (1K2AAAAAAAW2)
    • (if you are not yet subscribed) you may subscribe yourself to the [ ontology-summit ] listserv, by sending a blank email to <ontology-summit-join [at] ontolog.cim3.net> from your subscribing email address, and then follow the instructions you receive back from the mailing list system.     (1K2AAAAAAAW3)

Audio Recording of this Session     (1L)

Additional Resources     (1M)


For the record ...     (1M4)

How To Join (while the session is in progress)     (1N)


  • Shared-screen support (VNC session), if applicable, will be started 5 minutes before the call at: http://vnc2.cim3.net:5800/     (1N4E)
    • view-only password: "ontolog"     (1N4E1)
    • if you plan to be logging into this shared-screen option (which the speaker may be navigating), and you are not familiar with the process, please try to call in 5 minutes before the start of the session so that we can work out the connection logistics. Help on this will generally not be available once the presentation starts.     (1N4E2)
    • people behind corporate firewalls may have difficulty accessing this. If that is the case, please download the slides above and running them locally. The speaker(s) will prompt you to advance the slides during the talk.     (1N4E3)
  • Discussions and Q & A:     (1N4F)
    • (Unless the conference host has already muted everyone) Please mute your phone, by pressing "*2" on your phone keypad, when a presentation is in progress. To un-mute, press "*3"     (1N4F1)
    • You can type in your questions or comments through the browser based chat session by:     (1N4F2)
      • instructions: once you got access to the page, click on the "settings" button, and identify yourself (by modifying the Name field). 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.     (1N4F3A)
    • (when everyone is muted) If you want to speak or have questions or remarks to make, please "raise your hand (virtually)" by click on the "hand button" (lower right) on the chat session page. You may speak when acknowledged by the speaker or the session moderator (again, press "*3" on your phone to unmute). 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 "*2" on your phone to mute yourself after you are done speaking.)     (1N4F4)
    • 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) ontolog_20110303@soaphub.org ... Handy for mobile devices!     (1N4F5)
  • Please note that this session will be recorded, and the audio archive is expected to be made available as open content to our community membership and the public at-large under our prevailing open IPR policy.     (1N4J)

Attendees     (1O)


This page has been migrated from the OntologWiki - Click here for original page     (1O5)