ppy/OntologySummit2013_Symposium_chat-transcript_edited_20130503c.txt ------ Chat transcript from room: summit_20130503 2013-05-03 GMT-08:00 [PDT] ------ [05:45] Symposium (SteveRay): = Welcome to Day 2 of the Ontology Summit Symposium 2013 = [05:26] anonymous morphed into BruceBray [06:01] SylviaSpengler: Must better sound today...thank you [06:02] SteveRay: Please move over to the May 3rd chat room. Just change the URL to end in 0503 instead of 0502. The rest of us are over there. [06:04] SylviaSpengler: sorry wasn't paying attention...thanks Steve [06:04] SteveRay: See you on the other side! [06:00] == OlivierBodenreider presenting Keynote-3 [06:05] SylviaSpengler: could you remind the speaker about saying changed slide or its title.. [06:06] Symposium (SteveRay): Currently on slide 5 [06:22] GaryBergCross: It is interesting to see that under organizing principles there is no hierarchical structure among the major concepts. [06:25] anonymous morphed into FrankOlken [06:26] HansPolzer: Interesting that the assessment approach doesn't include anything about the purpose of the assessment or the context/perspective from which the assessment is being made [06:28] TerryLongstreth: Hans - sometimes it does, abstractly, as in chart 26 [06:40] GaryBergCross: Anatomy point seemed like a problem identifying parts vs. object in region of pelvis. [06:41] SteveRay: [ref. slide #49] +1 to that. [06:41] GaryBergCross: [ref. slide #50] @MikeDean Another use to test our Path ODP!! [06:43] SteveRay: [ref. slide #51] Intrinsic evaluation! [06:45] Symposium (SteveRay): slide#55 [06:46] GaryBergCross: In a way these quality measures set Requirements" for the ontology [06:47] SteveRay: I think the heading should be Diabetic patients rather than Pediatric patients [06:48] SylviaSpengler: not according to the ages [06:48] BruceBray: in this case it is diabetic pediatric patients [06:49] SteveRay: Ah [06:49] SylviaSpengler: agree [06:49] SteveRay: Ah [06:49] GaryBergCross: @ Steve Does age 5-17 set this as pediatric? [06:49] BruceBray: yes [06:51] BruceBray: in this application the age specifies the denominator or cohort specification for the measure (ie. eliminate very young children and adults) [06:55] GaryBergCross: For me OlivierBodenreider's talk brings some of the quality discussion into focus via real examples of what had been discussed generally. [06:55] TerryLongstreth: Thought experiment - does updating their ontology periodically create new ontologies every time? Are the ramifications of all changes communicated to all dependent applications/users/uses? [06:55] GaryBergCross: @Terry maybe it is just versioning. [06:56] SteveRay: @Gary: This underscores my belief that most benefit can be derived from the simpler problems (i.e. "intrinsic" error checking), much of which can be automated. [06:57] MikeBennett: Change management in ontologies sounds like another very important aspect of the ontology development lifecycle. We've considered the lifecycle for the initial development of the thing, but with any engineering artifact, changes into subsequent versions is also part of the process (and arguably the rationale for having a good process to begin with). [06:57] BruceBray: @gary, I agree - it would be very interesting to apply some of the tools/methods discussed in this summit to the biomed domain [06:57] SteveRay: The change management problem looks identical to the problem handled in code development. [06:57] TerryLongstreth: @Gary - whatever it's called, dependencies against the prior version are in an unknown state after the changes are applied [06:57] GaryBergCross: @MikeBennett Since ontology development is never complete the versioning problem is intrinsic to the effort. [06:58] AmandaVizedom: Very important thing to note, IMHO: The need to support changing codes that Olivier just mentioned (and language use, in other cases) is among the reasons to observe discipline in representing concept/entity structure centrally, and map to different code systems (i would include NL terminologies among code systems, here). Updating mappings is very different that updating concept/entity structure or names, and updating mappings is a much more manageable process. [06:59] GaryBergCross: @Terry The relations need not be unknown, but in practice this quality may not be realized because of lack of resources and standard methods and tools. [07:01] DougFoxvog: Change management should include policies for handling changed codes. Those removed have recommendations for replacement codes (sometimes one replacement, sometimes more). KBs need to be scanned for the changed codes and the recommendations for changes for specific assertions need to be generated. [07:01] anonymous morphed into Misha [07:02] AmandaVizedom: Another suggestion to support ease of updates *and* preserve interpretability of older data: If code mapping include explicit naming of the code system *and version*, mapping to new code systems can be added in new versions of the ontology in a way that facilitates detection when out of date, tracking, etc. [07:06] GaryBergCross: I wonder if anything analogous to regression testing makes sense and/or is done? [07:07] AmandaVizedom: @OlivierBodenreider - thank you for presenting this. It's an excellent example of good, integrated evaluation practices that have concretely improved specific ontologies. The detailed look is wonderful to have. [07:08] == The OntologySummit2013_Communique - FabianNeuhaus presenting [07:08] AliHashemi / AmandaVizedom: see the Communique at: https://docs.google.com/document/d/1gaQ8DIdg20AGJnlaCpQ7ebwJH_L4c4tK5RH1wyMHQig/edit?usp=sharing [07:09] DougFoxvog: The date of assertions can provide a maximum version of the code set(s) used. A KB that tracks the changes to codes could date the last time each code's meaning/properties is changed. [07:10] AmandaVizedom: Remote participants: you may need to refresh the session page to see the link to the slides. [07:12] anonymous morphed into ClarePaul [07:14] AmandaVizedom: We are now on slide #7 [07:16] GaryBergCross: Should craftsmanship include good analysis as well as design? [07:19] MatthewWest: I think good analysis is what fidelity is about. [07:22] GaryBergCross: @Matthew Yes, I can see that. But you might also have to design analysis products to engage is discussion with domain experts such as conceptual viz. [07:25] MatthewWest: @Gary: I agree, but Craftsmanship is not a lifecycle phase, but something that applies to artifacts of different lifecycle phases, and so is fidelity. So I would still consider that good analysis is detected by a test of fidelity, where ever fidelity is important in the various artifacts produced. [07:26] MikeBennett: @Gary the question there is what method is used for presenting content to domain experts - do you design a one-off throwaway viz to show them, do you design an interface which shows all or some of the actual model formalisms directly (do you limit what is expressed in the ontology to what can be presented to this audience). Or do you use ACE etc. So once the answer to this requirement is known, there will be design considerations for how that is handled. [07:24] AmandaVizedom: Please keep in mind, also, that this presentation is necessarily just an outline or summary of the points in the Communique. [07:25] GaryBergCross: @Amanda Will the slides be identified and available as background into to help interpret the communique? [07:26] DougFoxvog: @Gary: Everything in the slides is already in the communique. [07:27] GaryBergCross: @Doug yes, but this is like an outline that gives an overview and some focus and some browsers might look here first and afterwards too. [07:26] BobSchloss: Two "artifacts" that Fabian doesn't call out in his slides are: Example Queries (first expressed in natural language, later expressed in some semi-formal or precise query language), (perhaps this is one case of a capability question), and Illustrative Sample Instances. My experience is that we realize that our ontology has gaps or makes it too difficult to code and execute the most critical queries is discovered when we use these against our intermediate designs. [07:28] BobSchloss: I think Matthew and I are both thinking about how tests of ontology adequacy are performed. Mike points out that in many cases, a valid test involves domain experts who will have no willingness or capacity to look at the representation of the ontology design which professional ontologies or data modelers use. [07:28] MatthewWest: @Bob: Yes that is covered in the communique, at least at a high level. That's one of the ways that requirements are determined. [07:27] AmandaVizedom: In the Communique, we had many, many hard choices to make regarding how much detail to go into where. And even after we cut much that could have been said, the document is still long... We tried to select with an eye toward making a strong and clear statement, and advocating effectively in the direction in which consensus and understanding emerged... [07:29] DougFoxvog: @Amanda: I wouldn't say that the document is too long. It covers a lot of material. If it were arbitrarily limited to 10 pages, it would be incomplete. [07:30] BobSchloss: @Amanda - I do not think the communique fails just because I would have written somethings more specifically -- I'm just sharing my own experience with the people who are at this meeting. (Yes, I know I'm bad for not writing things for you and Fabian weeks ago -- work is incredibly busy!!) [07:31] AmandaVizedom: We hope that the Communique will encourage people to look further into the supporting resources ... [07:31] MikeBennett: If the communique is anything like what Fabian is summarizing here, it sounds to me like they have got to the heart of the important issues that have come out of these discussions - not an easy task. [07:31] AmandaVizedom: *and* those resources should and will continue to evolve after today. [07:32] MikeBennett: Perhaps the communique can be followed up by a more referenced / academic white paper covering this material - I think the specific things that have come out of this would merit that (but it sounds like a lot of work for someone). [07:34] AmandaVizedom: To ToddSchneider's point (that there are tools available to augment the development of requirements): We are not talking about generic software requirements tracking tools. We are talking about tools that specifically are integrated into ontology development, management, and evaluation tools. [07:35] BobSchloss: It would be an interesting exercise (but not something I am signing up to do) to try to document this approach in some tool such as Rational Method Composer - http://en.wikipedia.org/wiki/IBM_Rational_Method_Composer https://jazz.net/library/article/1092/ Then tooling to support an ontology development and release process, at least that based on Eclipse, could actually have a sidebar with "What to do next?" for novice ontology developers. I think about this because I am not sure that all commercial organizations have the budget to hire experienced ontology designers so that their own staff can apprentice with these experienced and skilled people. [07:37] AmandaVizedom: To date: outside of a few, proprietary examples (e.g. Cyc), there are no true Integrated Development Environments for ontology, as there are for software development. [07:34] SteveRay: All footnotes after #9 seem to have disappeared in the Google Doc. [07:38] AmandaVizedom: @Steve: I still have them showing... [07:39] AmandaVizedom: All: If you do not see all footnotes, you should be aware that we have noticed a bit of bugginess in gdocs around this... [07:41] AmandaVizedom: Occasionally the footnotes seem to hide from people, possibly when using browsers other than chrome. [07:47] SteveRay: @Amanda: Indeed footnotes don't seem to work in Safari but do work on my machine in Chrome. [07:42] AmandaVizedom: We can produce a pdf download snapshot of it if needed. [07:46] PeterYim: @Amanda & Fabian - I have downloaded a version (from your google-doc) and is versioning that as v1.0.0 dated 2013.05.03-10:30 which people are endorsing ... will post that pdf on the wiki later too [07:51] AmandaVizedom: Thank you, Peter. [07:58] GaryBergCross: I would just add to this discussion of "ontologies are part of systems" the idea of ontologies as reference models (mentioned by Leo) and the type of use of ontology in Biomed noted by Olivier. These seem one step removed from Apps by can be used there. [08:05] anonymous morphed into DanCarey [08:06] FrankOlken: If Cyc has had an integrated ontology development environment for 15 years, then it would be difficult for NSF/CISE/III to fund such a project as research - unless a case could be made that the proposed ontology development environment was a substantial improvement over prior art. It may be possible to fund an ODE via either Computer Research Infrastructure program, or via the ACI (Advanced Computer Infrastructure) division's software tool development programs. [08:31] AmandaVizedom: @Frank: note that the Cyc environment is not in public domain, and is customized for use with the CycL language, Cyc inference engine, and supporting infrastructure test harness, test- and quality-related ontology, automated bookkeeping, inference tracing... To the best of my knowledge (which I admit could be out of date), it is not usable outside of those conditions. For example, it could not be used on ontologies in OWL or CLIF. And key pieces are not part of OpenCyc (I believe because including it would allow too much, and too easy, reverse engineering of the related proprietary Cyc technology) [08:38] AmandaVizedom: (continuing prior comment: so, it is a huge advantage for projects using licensed, full Cyc. But unless the broader community using standard languages all start buying Cyc licenses and working natively in CycL, it does not help the general ontology community, or interoperability-oriented community using ontologies. [08:38] AmandaVizedom: ) [08:41] AliHashemi: yes! [08:43] DougFoxvog: A significant subset of the Cyc environment *is* available through earlier versions of OpenCyc (through 2.0). OpenCyc 4.0 makes it difficult to access the Cyc environment. [08:34] PeterYim: == session resuming (after the break) ... [08:36] PeterYim: == the Hackathon-Clinics overview on behalf of the co-champions of this group of H-C Activities [08:40] PeterYim: === KenBaclawski: comments and suggestions for the Hackathon-Clinics activities? ... [08:41] PeterYim: [consensus] we should do the Hackathons & Clinics again [08:42] AliHashemi: the scope is sometimes too much for one session [08:44] anonymous morphed into EvanWallace [08:45] AmandaVizedom: My suggestion: (1) start earlier, (2) allow for the possibility of multiple iterations. For example, 1 day intensive work, then a few weeks later, after time to think about it, another 1 day intensive. [08:46] AmandaVizedom: Another thought: allow for the possibility of some *entirely* asynchronous (any time of day) projects, within a limited time span (say a weekend). I imagine that this could work well for highly parallelizable projects with a good idea of what they want to get done. [08:46] anonymous morphed into DjokoSayogo [08:47] TerryLongstreth: I suggest we separate hackathons from the summit, and have them independently scheduled and planned [08:52] AmandaVizedom: There is, though, a lot of benefit from the flow of ideas between the summit activities and the hackathons. Benefit for both, I think, from the relatedness of focus. [08:57] TerryLongstreth: Related but different topic - followup on this summit - add a hotlink in footnotes to Communique for posthoc comments, observations, activities - in essence a 2013Summit blog - which could also be a place for FOLLOWUP hackathons. [08:50] PeterYim: DougFoxvog noted that we should not have scheduled the activities over the Easter Weekend ... and the point was well taken [08:49] FrankOlken: Catholic Easter or Orthodox Easter? [08:57] DougFoxvog: @Frank: commercial Easter. The one with Easter bunnies, eggs, and family get-togethers (in the US). FWIW, Orthodox Easter has not been interfered with (which may be important for the Russians) -- it occurs this coming weekend. [08:51] anonymous morphed into NicolauDePaula [08:54] FrankOlken: I am opposed to pushing the OntologySummit into June. It would run into Semantic Technology Conference in SF or possibly SIGMOD/PODS conference. [08:54] == JoelBender was unable to join us to present on the HC-01 project - SteveRay made a few remarks as a participant in that activity [08:56] anonymous morphed into JamesMichaelis [08:59] FrankOlken: Steve, I agree that ontologies would help tighten up many standards. I would point to work of HL7 on HL7 version 3 which is based on a reference information model from which HL7 messages are derived as views on the underlying Reference Information Model. [09:01] MikeDean: I tried to comment on BACnet, but apparently only introduced echo [09:02] SteveRay: Sorry Mike [09:02] MikeDean: My main point is that ASN.1 describes messages, so there was a need to represent messages as well as the domain model [09:00] == HC-02 The General Ontology Evaluation Framework (GOEF) & the I-Choose Use Case - JamesMichaelis presenting (remotely) [09:03] BobbinTeegarden11 morphed into BobbinTeegarden [09:04] anonymous morphed into JerrySmith [09:05] AmandaVizedom: @James, could you speak a tiny bit slower? It's a bit hard to understand at times. [09:12] anonymous morphed into DanCarey [09:15] anonymous morphed into KenBaclawski [09:22] GaryBergCross: It is interesting to note in passing that they cite 3 different criteria (from our 5) to evaluate the ontology - Correctness, Completeness, & Utility. [09:27] AmandaVizedom: @Gary - just to note, Joanne did present on GOEF during one of the summit sessions. So the characteristics/requirements/etc mentioned as part of the GOEF framework were among the many that we looked at, along with those from OntoQA, OQuaRE, OOPS!, other presentations, and prior literature, that fed into the ongoing summit discussions -- which discussions ultimately led to the clustering of in-scope potential requirements into the focal 5 we ended up with. [09:20] == HC-03 Evaluation of OOPS! OQuaRE and OntoQA for FIBO Ontologies - MikeBennett presenting [09:32] anonymous morphed into HansPolzer [09:34] BobSchloss: Listening to Mike talk about grounding FIBO concepts in legal, financial, regulatory concepts, I am reminded that in the Intelligent Semantic Model Palette (for Smarter Cities) that I have worked on with RosarioUcedaSosa and others, we adopted conventions to use annotations so that for any concept in ISMP (which internally we call SCRIBE), its complete or partial heritage or partial mapping to concepts in pre-existing representations, vocabularies or data communication standards (such as Common Alerting Protocol, National Information Exchange Model, DATEX II for transportation) is always specified. We think this may allow software developers familiar with these standards to find the matching concepts in the ISMP ontology. [09:40] GaryBergCross: @Amanda Is utility reflected in your 5 categories? deployability seems closer, but is not quite it. I searched and the word 'utility' is not in the document. [09:41] AmandaVizedom: Gary, it is reflected, especially in fitness, but may require some unpacking of implications to see... [09:43] AmandaVizedom: This comes under the heading of things we talked quite a lot about, that provided important content to the Communique, but that may have a non-obvious terminological alignment... [09:44] TerryLongstreth: @Gary - I'm curious as to how one would measure utility empirically. [09:45] GaryBergCross: @Amanda I can see that a bit, but the def doesn't easily lead me to think of utility which is derived from it- Does the ontology accurately represent its domain? (Fidelity) [09:45] MatthewWest: @Gary: Utility is really a function of meeting requirements. If it isn't then it means you did not get your requirements right. [09:45] AmandaVizedom: (I have particular hopes for the future work along the Ontology of Ontology Evaluation lines to capture such alignments and relationships, including the work that MikeBennett just talked about.) [09:45] AmandaVizedom: (more on that after lunch) [09:47] Symposium (SteveRay): == session adjourned for lunch ... reconvening in about 50 minutes [09:48] Symposium (SteveRay): == all remote participants will need to dial-in to the conference bridge again, after this lunch break. (Please dial-in a couple of minutes early if you can.) [10:44] anonymous morphed into DanCarey [10:46] PeterYim: please dial back into the conference bridge [10:51] Symposium (SteveRay): == HC-04 OntologySummit2013 Content Hack: Leveraging Semantics on OntologPSMW - KenBaclawski presenting [11:05] == HC-05 Hackathon: Ontology of Ontology Evaluation - AmandaVizedom presenting [11:09] AmandaVizedom: Observation: that indirection in the ontology is actually a typical sign that what you have is really not a standard ontology but a subject hierarchy / classification system. You can either understand the subsumption relationship to be broader/narrower in the subject-headings sense, or treat all of the class names as containing an elision of "information about..." (or "pages about...," in this case. [11:11] MikeBennett: @Amanda that came up in some of the classification research we did with the folks at UBC - for instance with books, French Grammar is not a kind of French, but a French Grammar Book is a kind of Book about French. The "Thing" is the book / article etc. not the subject. Sounds like like is applicable to wiki as well. [11:16] GaryBergCross: I'd like to a see a Hackathon on converting some existing subject classifications into a useful ontology! [11:30] GaryBergCross: @Amanda This work is a good opportunity to integrate with some IT methods and show Process diagram as well as an Object model both of which then get represented in an Ontology. [11:33] AmandaVizedom: @Gary, I like that idea. [11:16] anonymous morphed into DanCarey [11:25] HamizahHamka: Amanda, did you consider the meta ontology proposed by AldoGangemi? [11:27] HamizahHamka: Oqual [11:28] HamizahHamka: http://km.aifb.kit.edu/ws/eon2006/eon2006gangemietal.pdf [11:29] AnatolyLevenchuk: http://ontolog.cim3.net/forum/ontology-summit/2013-04/msg00038.html [11:32] AmandaVizedom: @Hamizah, we did not look at that one during the hackathon day, though we do have it in the collection of relevant references. Thanks for the reminder. [11:33] MikeBennett: @Gary in FIBO we used Occurrent Things partition to create process primitives and so draw process flows that were also ontologies - a bit primitive, since then DennisWisnosky has worked on something in OMG which maps BPMN to OWL. [11:37] GaryBergCross: @MikeBennett Yes , I was thinking of a simple Endurant / Occurent distinction with these on a top level of the diagram so below things are sub-types of these 2. [11:43] BobbinTeegarden: @MikeBennett BPMN? But wouldn't a UML activity diagram be a richer expression of the graph structure needed? [11:44] AmandaVizedom: @Gary, @Mike there are a number of fairly well-established ways to ontologize the processes and process relationships, including scripts, subprocess hierarchies, specification of participants and the role they play, etc.. Any of those could be applied in formally and shown in a diagram which is itself quite understandable. I may have misunderstood Gary's first point; I took it too me about the usefulness of having visual conventions (e.g., color, shape) for showing the process and event things differently from, say, the actors and information objects. [11:45] BobbinTeegarden: @Amanda Shape, not color, my vote. [11:48] AmandaVizedom: @Bobbin - certainly color runs up against limits and perception variations more quickly! [11:49] MikeBennett: @Amanda I'd like to look into that some more. I see potential for conceptual modeling tooling which goes beyond ontology to business process, business rules etc., with everything tied in to 1st order logic, math etc. - I like the idea of different appearances. [11:49] GaryBergCross: @Amanda You understood what I was saying, which was to use a color or shape code at this level, but that could be later formalized a bit by linking to DUL concepts. [11:50] MikeBennett: On appearances, there was a chap who presented at the last OMG meetings, on taking a more scientific approach to appearances of things - tore into UML and the rest :) [11:50] MikeBennett: (for instance, use color as a way to help the eye differentiate things but not as primary distinction which would be e.g. shape). [11:53] AmandaVizedom: @Bobbin - but good visualization and use of visual metaphors are areas in which I tend to rely on the strengths and expertise of others -- I recognize them as important, and also make a practice of attending to whether an intended audience (say, experts reviewing something) can easily and correctly understand some presentation, or whether it needs to be modified for them. In that (common) case, though, I often look for tools and/or metaphors that audience likes already, or find someone who has more UI / visualization expertise to help solve it. [11:53] GaryBergCross: @MikeBennett Yes the older convention seemed to be sq for continuent-objects and ovals for process-occurents. [11:54] AmandaVizedom: @Gary, Mike - increasingly crowded convention-space. Shape uses in different diagramming practices can conflict. [11:55] MikeBennett: @Gary is that published somewhere? We have the opportunity now to rethink and improve the business facing visuals in our work. [11:56] SteveRay: I would recommend turning to UI research such as the work of Ben Schneiderman for some good principles. [11:56] BobbinTeegarden: @Amanda Agree that human sensory reaction important. Thought: color doesn't print well, turns out it's important for model usage, I think. [11:57] AmandaVizedom: @Mike, during the USAF project, we had that specific issue: bridging BPMN and OWL. Well, not only that; representing BPMN in OWL was useful but didn't solve our whole problem. We *really* wanted to bridge the work of Enterprise Architects and that of Ontologists, so that we could have more continuity &integration of work and reuse of information gathering. [11:58] GaryBergCross: @MikeBennett Data Flow Diagrams The formal, structured analysis approach employs the data-flow diagram (DFD) to assist in the functional decomposition process. I learned structured analysis techniques from DeMarco [7], and those techniques are representative of present conventions. To summarize, DFD's are comprised of four components: External interactors are represented by a rectangle. Data stores are represented by an open rectangle (2 or 3 sides). Processes are represented by any rounded object (a circle, oval, or square with rounded corners). A DFD process may represent system function at one of various levels, atomic through aggregate. Data flows are represented by arrows, with labels indicating their content. http://www.umsl.edu/~sauterv/analysis/dfd/DiagrammingMethods.html [11:58] MikeBennett: Turns out that UI chap is right here in Maryland http://www.cs.umd.edu/~ben/ [11:58] MikeBennett: @Gary, ah yes, I remember those. [11:59] AmandaVizedom: E.g. EA often bottoms out at the exchange of information objects (e.g., reports, data artifacts). It doesn't go into the information content of those objects or that exchange. That's where the ontologists on that project were picking up... [12:00] MikeBennett: @Amanda what I'd want to see is something where you draw the process flow in front of the users, so it looks exactly like a Visio swimlanes diagram, but behind the scenes each Process, Process Activity, Process Event etc. is given its 'is a' relationship to the kind of occurrent thing it is a kind of. Then in drawing the process you draw the process ontology. [12:00] BobSchloss: If changes are being made to OntoHub, would the team consider adding social media facilities -- such as where visitors can participate in a discussion forum, with questions, answers, vetted answers-to-questions (FAQ), around each ontology registered with the Hub, as well as perhaps discussion of the performance of different reasoners in different configurations with different databases with the ontology, etc. [12:00] GaryBergCross: @Amanda I guess, theoretically the Data Ref Models of EA were proposed to cover the "content". [12:00] SteveRay: @Mike: Indeed, he's local. As I recall, for example, he invented the notion of TreeMaps and space-filling maps, which are now very popular. [12:01] AmandaVizedom: And, of course, the processes provided context that was implicit in the data, and which we wanted to make explicit in the ontology. Plus, some of the content was about processes! [12:02] BobbinTeegarden: @Gary @MikeBennett yes, we have used Data Flow Diagrams for initial (unified domain) modeling, then broken it into UML process/structure(class) because of the need to get to code. The DFD model is a good place to start with 'active ontologies', IMHO. [12:02] SteveRay: Golden Rules of Interface Design: http://faculty.washington.edu/jtenenbg/courses/360/f04/sessions/schneidermanGoldenRules.html [12:02] GaryBergCross: On yes Ben Shneiderman. A guru in the field. [12:03] anonymous morphed into PeteNielsen [12:04] MikeBennett: @Bobbin the weird thing is that "pure" UML would not let you mix activity and class diagrams, yet you need to talk about inputs / resources and products of process activities (the Sparx EA tool lets you do hybrid diagrams for that). Vital for e.g. securities issuance process and related reference data elements. [12:04] AmandaVizedom: @Mike - Robert Kahlert (Cycorp) and I envisioned and prototyped just such a tool under the RKF project (circa 2003?). This in response to working with domain experts & trying to enable them to directly create terrain analysis ontology. There's a paper draft from that kicking around somewhere. [11:25] == HC-06 ISO 15926 Reference Data Validation - AnatolyLevenchuk presenting (remotely) [11:45] == HC-07 Ontohub-OOR-OOPS! Integration - TillMossakowski presenting (remotely) [12:06] BobSchloss: On Slide 13, in the table, row 3, column 2, there is a typo (it says langauge.id instead of language.id) [12:07] MikeBennett: This Federation API work would also be of interest in the OMG SIMF initiative (semantics of information models for federation), I think. (current slides) [12:07] AmandaVizedom: Motivator was that the experts could easily call up and articulate their expert knowledge while performing, simulating, or describing their terrain analysis processes, but not outside of that. Without the process context, the pieces either made no sense to them or seemed to be show-stoppingly subject to misunderstanding. [12:10] BobbinTeegarden: @MikeBennett MagicDraw let's you mix class with process in activity diagrams, just fyi, as I use it for those 'DFD' diagrams. [12:15] == MichaelGruninger announced that the winner of First IAOA Best OntologySummit Hackathon-Clinic Prize is the HC-03 team that worked on "Evaluation of OOPS! OQuaRE and OntoQA for FIBO Ontologies" [12:16] AmandaVizedom: Congratulations to the HC-03 team! [12:20] MikeBennett: Thank you! Here's a big shout out to AstridSuqueRamos, MariaPovedaVillalon, SamirTartir and also the rest of the team - Simon, Amanda, Peter, the FIBO folks (Jacobus Geluk, Derek LaSalle, Max Gillmore, Kevin Tyson and others), all of whom put in valuable ideas and perspectives to the event. [12:25] == MichaelGruninger and MatthewWest: Discussing Next Steps and Conclusion [12:25] GaryBergCross: (ref. the comment on less people coming to the symposium) Contributing Reason Sequestation [12:36] GaryBergCross: Suggestion Have a half day workshop preceding the Symposium for newbies who might be shy about attending w/o more knowledge. [12:30] JerrySmith: Peter - How do people 'sign' the Communique electronically? [12:36] PeterYim: @Jerry & ALL - for those who are online (i.e. now, virtually), you can just type in "I endorse the communique") for those who are not around, we will be posting a solicitation for endorsement to the [ontology-summit] mailing list shortly!) - ref. http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2013_Communique [12:41] MikeBennett: We would expect to build on the HC-03 hackathon work by building on the Ontology of Ontology Evaluation as a formal basis for structuring this work. [12:41] EvanWallace: @MikeBennett Some of the folks involved with Ontohub were initially involved in SIMF submission development. But I haven't noticed much involvement from those people in SIMF in recent months. I can think of many different reasons for that. SIMF was a great idea, but I doubt it will live up to its potential (if it results in anything at all). [12:44] MikeBennett: @Evan I'd forgotten there was some overlap. I'd like to have been more involved but we're an OMG Domain Mamber and so no eligible to participate. Been thinking of some deep meta-metamodel ideas, which reflects a small part of what I saw in those slides today. [12:42] anonymous morphed into DanCarey [12:45] FrankOlken: Lots of distortion on the Skype audio channel now. [12:46] == MichaelGruninger: join us on 23-May-2013 Thu (normal virtual session time; was 16-May when Michael announced this, date subsequently changed) for the OntologySummit2013 Post Mortem session [12:46] == MichaelGruninger: participants welcome to stay for the "Birds of a Feather (BOF) meeting about follow-up work on the "Ontology of Ontology Evaluation" (HC-05) work [12:47] FrankOlken: Maybe.... [12:47] AliHashemi: briefly though. [12:47] BobbinTeegarden: Yes, staying. [12:50] JerrySmith: CONGRATULATIONS! Each Summit gets better and better. The outstanding success of this one tops all of the others. Great job all!!! [12:51] vnc: == the OntologySummit2013 Symposium is now adjourned [12:53] List of attendees: AliHashemi, AmandaVizedom, AnatolyLevenchuk, AstridDuqueRamos, BobSchloss, BobbinTeegarden, BruceBray, ChristianHempelmann, ClarePaul, DanCarey, DjokoSayogo, DougFoxvog, EvanWallace, FrankOlken, GaryBergCross, HamizahHamka, HansPolzer, JamesMichaelis, JerrySmith, KenBaclawski, LaVernPritchard, MatthewWest, MeganKatsumi, MikeBennett, MikeDean, Misha , NicolauDePaula, OliverKutz, PavithraKenjige, PeteNielsen, PeterYim, SteveRay, SylviaSpengler, Symposium (SteveRay), TerryLongstreth, TillMossakowski, anonymous, anonymous1, vnc, vnc2 [12:52] vnc: === we are resuming in 8 minutes for the informal BOF on "Ontology of Ontology Evaluation" session [12:52] AmandaVizedom: 10 minute break, then Birds of a Feather session re: continuing work on ontology of ontology evaluation [13:03] PeterYim: == Birds of a Feather (BOF) meeting re: continuing work on ontology of ontology evaluation ... is now in session [13:06] AliHashemi: Ali committing to provide an English description of the informal model. [13:06] AliHashemi: Make the Hackathon page or a new Ontology of Ontology Evaluation page as centralized starting point. [13:06] AliHashemi: ? [13:07] MikeDean: I don't think you can hear me. I'm interest in continuing to work on and update the OWL formalization. [13:07] MikeDean: s/interest/interested/ [13:13] AmandaVizedom: subcollections of https://www.zotero.org/groups/ontologysummit2013/items/collectionKey/PVGF24A6 [13:13] PeterYim: the HC-05: "Ontology of Ontology Evaluation" Project Homepage is at: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2013_Hackathon_Clinics_OntologyOfOntologyEvaluation [13:14] PeterYim: shared-files (artifacts) are under: http://ontolog.cim3.net/file/work/OntologySummit2013/Hackathon-Clinics/HC-05_Ontology-of-OntologyEvaluation/ [13:15] AmandaVizedom: I am interested in continuing to grapple the graphical informal model. The final informal model will be submitted for team review, but in between I would love help, tool suggestions, etc. [13:17] vnc: the shadowed HC-05 project homepage at the OntologPSMW - http://ontolog-02.cim3.net/wiki/OntologySummit2013_Hackathon_Clinics_OntologyOfOntologyEvaluation [13:19] AmandaVizedom: Vote yes if you want to be involved and included in communication about continuing activity on this project. Then email me and tell me what email to use for that. (firstname.lastname@gmail.com) (1) Yes (2) No This is a single choice vote. [13:20] AliHashemi: (1) [13:20] AliHashemi: AliHashemi voted for: Yes [13:20] MikeBennett: MikeBennett voted for: Yes [13:21] DougFoxvog1: Yes [13:21] DougFoxvog1 morphed into DougFoxvog [13:21] PeterYim: PeterYim voted for: Yes [13:21] MikeDean: MikeDean voted for: Yes - mdean [at] bbn.com [13:22] GaryBergCross: GaryBergCross voted for: Yes - gbergcross [at] gmail.com [13:22] MikeBennett: Use mikehypercube [at] gmail.com [13:23] AliHashemi: http://arxiv.org/abs/1304.7224 [13:23] FrankOlken: There is a lot of work going on in the provenance community. Some of it in W3C working group. [13:24] KenBaclawski1: KenBaclawski1 voted for: Yes - I can help with ICOM at least [13:24] GaryBergCross: Qood grid: A metaontology-based framework for ontology evaluation and selection AldoGangemi [13:24] PeterYim: I suggest we volunteer MichaelGruninger on work related to OMV (as he leads the metadata effort for OOR too) [13:25] AliHashemi: http://www.ccs.neu.edu/home/kenb/ontologies/ucdo.owl [13:25] AliHashemi: Use Case Description Ontology - http://ontolog.cim3.net/cgi-bin/wiki.pl?OpenOntologyRepository_UseCases [13:28] KenBaclawski1: We should also include the Use Case Description Ontology (UCDO) [13:29] TerryLongstreth: Time to go find a beer... [13:29] TerryLongstreth: Bye all; Thanks for a great Summit!! [13:30] AliHashemi: Great summit, congrats to all. Bye! [13:31] PeterYim: the has been marvelous! ... bye everyone ... talk to you all again on May-16, if not sooner! [13:31] AmandaVizedom: Thank you all! ------