chat-transcript_20100415c_unedited.txt PeterYim: Welcome to the OntologySummit2010 review and follow-up action planning session - Thu 15-Apr-2010 * Chair: Dr. SteveRay and Dr. FabianNeuhaus The goal of this session is to revisit the last Ontology Summit and plan ahead for the future. We will discuss what worked and what did not work during the 3~4 months of OntologySummit2010, and get ideas how to make next year's Ontology Summit even better. This meeting is also an opportunity to suggest topics for the next year's Ontology Summit. Further, this meeting gives us an opportunity to revisit the actions items listed in the end of the Communique and to identify members of the community who would be interested in participating in projects to address them. . Please refer to details on the session page at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2010_04_15 . EricLindahl: Are most people on VNC? anonymous morphed into KurtConrad PeterYim: VNC is just coming up ... there are 4 people on the conference and 4 in the chat-room (but slightly different people now) SteveRay: Hmm. Getting a busy signal on the phone number... SteveRay: 218-844-8060, right? anonymous1 morphed into doug foxvog anonymous morphed into Bobbin Teegarden PeterYim: For OntologySummit2010 - What went right? ... please type them in PeterYim: Full 3 month period ... from the Launch in Dec-2009 to the Symposium in Mar-2010 RexBrooks: I think the survey needed a bit more work to get more focused results. FabianNeuhaus: rex: which survey are you referring to? RexBrooks: I think that having the surveys was a great improvement, though. AmandaVizedom: I think surveys added a lot, *and* along Rex's line, would have added even more if planned and launched further ahead of time. ArturoSanchez: @SteveRay: I enjoyed very much the F2F meeting. We all had the opportunity to share our perspectives as part of collaboratively composing the communique. It was also great I was able to meet in person some of the folks I interacted with via conference calls before the F2F. AmandaVizedom: I think we improved outreach, *and* have a lot more room for improvement there. RexBrooks: I didn't take the real time delphi, but the other two needed to focus in on specific questions. Antony Galton: There was quite a lot of uncertainty at the beginning about the different track and what they meant - but I think we more or less sorted that out eventually. RexBrooks: @Amanda, I agree that working on the surveys perhaps a month ahead of when it started this year would give those creating the surveys more time to get feedback on focus. PeterYim: I guess we are doing "What needs improvements?" too now ... FabianNeuhaus: @ antony: I agree, this was very difficult this year to slice up the discussion in tracks. ArturoSanchez: @RexBrooks: please bear in mind some of us are not trained in preparing surveys ... Surveys are great tools for mining information from communities and it would be great if people with experience in preparing surveys could volunteer RexBrooks: @Arturo: I agree. I think that searching for contacts in survey management and perhaps market research would pay big benefits. EricLindahl: While I've loosely followed the development, it would be nice to have an ongoing developing 'Getting Started' artifact providing link lists and glossary/overview for people new to ontology development. E.g. http://www.google.com/search?q=ontology+training+site%3Acim3.net RexBrooks: We could trade benefits to the surveyors since ontological-semantic questions could be helpful to advancing the state of their art, especially if we can educate them about Linked Open Data for instance. RexBrooks: @Steve, that's why I suggested a benefit for the professional surveyors. EricLindahl: A concrete environment, ontology sandbox would be good. Rather just downloading Protege. Example texts gathered to act as ontology exemplars (like the wine ontology) anonymous2 morphed into JulitaBermejoAlonso anonymous1 morphed into PierreGrenon EricLindahl: What site will provide these materials? Are there copyright problems? ArturoSanchez: Did we hear an UK-style ambulance passing by? JulitaBermejoAlonso: @Arturo: French ambulance, sorry, I did not unmute the phone. Running late with my kids! ArturoSanchez: @Julita: ahh! No worries ... I think it is great we can hear street noises from far-away places in real time. De dónde eres, Julita? ArturoSanchez: @Julita: "De do'nde eres"? JulitaBermejoAlonso: @Arturo: Spanish, but living in France after the States. University in Madrid, though. A long story. EricLindahl: I wonder how closely the goal of this communique is being obviated, to some extent, by the evolution and adoption of modeling frameworks, like Eclipse Modeling Framework. EMF will likely be the de facto toolset for 'domain interoperation', one of the main factors for ontology development. ArturoSanchez: @Julita: nice meeting you EricLindahl: Or should I say, tree rewriting. AmandaVizedom: FYSA, as the committee knows, I submitted a proposal to present the summit findings at SemTech, but have not heard back and don't know whether that's likely to happen. I think that such presentations could offer an opportunity to do some more of the bridge-building I was talking about. I can also imagine working it in as part of a larger session or workshop involving folks from some of the stakeholder communities, tailored perhaps to other audiences. Do others have thoughts about particularly good venues for something like this? Next year's ISWC, for example? Other? anonymous1 morphed into PierreGrenon EricLindahl: It would seem that 'ontologists' learn something from the fact that Google is rather 'ontology hostile'. Tools have a way of treating ontology as a rewrite problem. Concept-net, BigTable, HBase, etc. PierreGrenon: It would be useful to have something between the communique and the futurist paper, perhaps minus the particulars account of the Delphi experiment. For example, in order to approach course production teams at the OU, I could use a write up that explains the findings of the summit and has enough context... AmandaVizedom: @Pierre - I agree, we should have this. A not-incidental side effect, if the SemTech proposal were accepted, would be that we'd have to create one! PeterYim: @Steve: ref. the whitepaper for "State of the Future 2010" ... (while that book is not an open publication) I will work with the Millennium Project people about allowing the open publication of our paper (on the Ontolog website, for example) EricLindahl: What happened to the ontology of measures? Measures seems prima facie for building working systems. e.g. http://jscience.org/api/javax/measure/unit/package-summary.html RexBrooks: @Arturo: I would be happy to help with a Service Interface to the Registry because I am active in the OASIS SOA Reference Model Technical Committee, the Reference Architecture Subcommittee and the efforts to align OASIS, Open Group and OMG SOA efforts. So I can socialize the Registry in those venues. PeterYim: @Eric - that project has now moved on and has morphed into the OASIS QUOMOS TC - see: http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=quomos SteveRay: @Peter: I think that would be a good thing to pursue, in the spirit of the Ontolog Forum. PeterYim: @Steve - agree totally PierreGrenon: @amanda -- re. submissions; perhaps some knowledge management and/or business inforatics conference? RexBrooks: @Amanda: I am speaking at SemTech, and I believe I can work a mention of the Ontology Summit and these various follow-up efforts as part of the context in which Semantic Technology tools are developing. RexBrooks: @Amanda: I don't think specifically mentioning the results is a good connection, unless you are actually presenting the results. EricLindahl: @PeterYim Thanks! @RexBrooks The market is moving more towards RESTful systems (which subsume many web services) which is design by *convention* NOT by *contract*. This is opposed to proving or even negotiating metadata by ontology PierreGrenon: @amanda -- re. propaganda paper happy to read any draft whenever they come if that helps RexBrooks: @Eric: A lot of RESTful applicatons are gaining traction but i wouldn't put all my eggs in any one basket. It's a horses for courses world. RexBrooks: In a world of emergency management and law enforcement information exchange by contract only will remain the case where necessary. EricLindahl: @RexBrooks Having just read a 'large' government RFP, it was pushing towards RESTful. EMF is essentially a RESTful technolgy. However, I do agree. EricLindahl: This opencourseware idea is excellent. RexBrooks: I suspect RESTful will find a receptive audience in the cloud computing world that is developing, and we need it in the portion of the information spectrum that allows public input and can aid immensely in emergency response. PierreGrenon: re. 'open source': nice if that happens, everybody should be encouraged to put out their material out. very useful from the community would be a public domain template curricular structure against which institution specific outputs may be checked EricLindahl: @RexBrooks I have some experience with HBase and Cassandra, which require little to *no* pre-designed schema. Need a new attribute for 1 specific row, add it. No ontology required. RexBrooks: @Arturo: Linked Open Data between registry participants could be very valuable. EricLindahl: I would like to see a code.google.com project where we can check in these artifacts, as Rex suggests. PeterYim: @Arturo ... please consider using OOR as part of your infrastructure for the "Ontology Education and Training Registry" initiative RexBrooks: @Peter: I agree, combining OOR with Linked Open Data could be extremely valuable. SteveRay: Can someone provide a reference to SCORM (?) or other metadata? EricLindahl: Perhaps I will take it upon my self in the May to create a code.google.com project (or equivalent) where I'll check in whatever artifacts come across the 'Ontolog' desk. RexBrooks: @Eric: I'd be happy to check it out and use whatever I can and suggest it to the groups I work with. ArturoSanchez: @RexBrooks: Thanks for the suggestion EricLindahl: @SteveRay the first 5 google links are fairly good. US government uses SCORM quite a bit (I've found Moodle a huge PITA, IMHO) ArturoSanchez: @PeterYim: thanks, Peter ... Noted! EricLindahl: by sandbox, what is Peter referring to? Not Protege? RexBrooks: @Eric: I missed that. There's a sandox in the wiki where people new to using the Ontolog wiki can learn how to use it, but he may have been referring to something more generic. PeterYim: @Eric - see: http://ontolog.cim3.net/cgi-bin/wiki.pl?OpenOntologyRepository ... the OOR sandbox is (currently) based on the NCBO BioPortal codebase - see: http://oor-01.cim3.net/ontologies anonymous3 morphed into dsoergel RexBrooks: @Peter: Ahh, yes! dsoergel: Metadata for educational materials dsoergel: *Learning objects (instructional materials): The Gateway to Educational Materials (GEM) www.thegateway.org/about/documentation/metadataElements Learning Technology Standards Committee of the IEEE: http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf IMS Global: IMS learning resource meta-data information model. (September 2001) www.imsproject.org/metadata/ Dublin Core Metadata Initiative. DCMI Education Working Group, http://dublincore.org/groups/education/ (not much concrete to see there) CRP Henri Tudor-CITI: Training Exchange Definition: TED. www.xml.org/xml/schema/8dbca03a/trainingExchangeDefinition.pdf (July 2002) PeterYim: @Eric - not Protege (which is an ontology development platform) BioPortal is more of a repository for users to "share" their ontologies (asfter they are developed) Antony Galton: I'm afraid I have to go now. Hope to see some of you at FOIS! PeterYim: Thank you Antony ... see you at FOIS! ArturoSanchez: @AntonyGalton: me too anonymous2 morphed into dsoergel EricLindahl: @PeterYim Thanks. I'm hoping for something like 'hg clone http://code.google.com/p/boot-ontolog/' I'll see what I can do. Thanks again. PierreGrenon: Could there be a notion that the programmes contemplated could perhaps become pilotes for the summit or the iaoa? This could drive the definition of programmes but also could serve as references when trying to achieve similar things in places that are less aware of ontology at the moment. EricLindahl: My suggestion: Tools & Technologies. This is where the 'rubber meets the road'. IOW where I work everyday. EricLindahl: I second 'OntologyFest' RexBrooks: I would suggest "Applications and Tools for Practical Ontology" or "Ontology in the Enterprise"? PeterYim: Now brainstorming on "suggestions for 2011 Ontology Summit" ... please make sure you document ALL suggestion into this chat-baord (as we will need to look through these suggestions again later in the year when we actually start organizing OntologySummit2011 EricLindahl: With EMF you can use the 'Ontology' right away. RexBrooks: I think Case Studies and a Hackathon at the face-to-face might be helpful. MichaelGruninger: Address the question -- how are the methodologies, tools, and environments for ontological engineering different from those for software engineering? EricLindahl: In support of Arturo, I suggest systems like QVT or ATL AmandaVizedom: @Rex: Good to know. There are a few others from the Summit who will be speaking as well, and if the proposal is accepted, the plan is to draw from that presences/support. It was a late submission, however, and I really have no idea whether it is likely to be selected. But I think it would be good to make such a presentation - between the communique and the gory details, as Pierre suggests. So, I think it's worth thinking about other gatherings of potentially interested people. AmandaVizedom: @Pierre - Count on being held to that! PeterYim: please propose some "Themes" ... worth of a *Summit* PeterYim: question we should ask ... by end this year ... what would be the *most* strategic issue that this field (of ontology) should be addressing? RexBrooks: My thinking is that it naturally follows training new ontologists, that we look at what they would actually be doing at work? PierreGrenon: theme: good ontology PeterYim: (3 lines above) I meant "worthy of a *Summit* EricLindahl: @PierreGrenon implied 'Ontology of Utility' ArturoSanchez: For the record. The two recommendations I voiced are: (1) Examples of systems for which the use of ontology technology was successful and otherwise; (2) Sessions similar to "design-fest" and "code-fest" from OOPSLA, which in our case would be hands-on sessions on the design/use/testing/integration/etc of ontologies. I think this can be combined with the Michael Gruniger's suggestion. PierreGrenon: @eric: not sure i follow AmandaVizedom: "Ontology Practices in Context" or something, meaning: collecting info on, and developing some collective understanding of, how ontology practices (both actual and best) vary with elements of context such as application type, developer community, user community, large goals (monetization vs. decision support, reuse vs. one-time-need)... EricLindahl: Good is a utility function. Applies to decision support, marketing etc. PeterYim: Fabian suggested the topic on "Ontology Modularity" RexBrooks: the problem with word "good" is the difference betwen useful (utility) versus effectiveness (value to user for achieving objectives). EricLindahl: To say 'good(Ontology) range [0..1)' (assuming it's unit). I prefer DS or rational RexBrooks: That's actually a valid question for an ontology of meanings. PeterYim: please also think of "involving another community" - like OntologySummit2009 when we got the ontology and the standards communities to get together AmandaVizedom: In a way, my suggestion could integrate Pierre's (good ontology) and Arturo's (what has worked where? & design-fest) and would have to incorporate draw from evaluation / standards. PierreGrenon: @fabian: i agree this is a topic that could cause troubles RexBrooks: @Peter: that's what I was aiming for, combining standards for ontology and standards that use ontologies as representations in the practical world: Case Studies. EricLindahl: I submit that Utility is a better and more tractable subject, than whether a given ontology is better than another. That presumes too much, IMHO. anonymous morphed into JulitaBermejoAlonso RexBrooks: @Eric: agreed! AmandaVizedom: @Peter - I also think the "practice in context" idea inherently brings in other communities. ArturoSanchez: @Peter-and-Michael: I like very much the idea! PierreGrenon: well it could be purpose / context oriented. No absolute norms of q PierreGrenon: but good for a purpose. We could then include application and tool developers EricLindahl: Specifically, perhaps, SizeOf is a better quality measure than Good. E.g. Equivalance partition on KolmogorovComplexity(Ontology) http://en.wikipedia.org/wiki/Kolmogorov_complexity MichaelGruninger: @Peter -- yes, bringing the ontological engineering and software engineering together -- that is in the direction that I was thinking ArturoSanchez: All, I need to go now. Good session. Thank you for the feedback offered! RexBrooks: @Michael & Peter: Me, too. PeterYim: something that is important for the two communities, as Michael suggested, bringing together the ontology community and the software engineering community EricLindahl: I disagree PierreGrenon: @amanda: yep, that sounds very good, but hard AmandaVizedom: To put my suggestion differently: We know that we have differences in views on Methodology. Can we discover relationships between methodologies used, application/community, and success? I think so, and it gives a way to make headway on quality & best practices without searching for universally agreed upon, context-free principles. PeterYim: other possibilities is to being together the "library science community and the ontology community" ... another one, "enterprise architecture and ontology" PeterYim: I meant "to bring together" MichaelGruninger: Ontological engineering can learn about methodologies and best practices from Software Engineering, while Software Engineering can learn about how software integration and model-directed architectures can be supported by the application of high quality ontologies RexBrooks: I think using Cast Studies: successful and not is very useful for engaging other communities like software engineering, but I wouldn't focus on SE alone. The domains that SE deals with are equally important as end users. RexBrooks: I meant Case Studies. AmandaVizedom: @Pierre: Does it? It seems moderately hard but doable to me, because we could again develop questions to elicit information to ground the discussion in. Harder to have nothing empirical, I think, and more likely to degrade into idealogical battles. EricLindahl: Library science is a KR problem, right? RexBrooks: I wouldn't focus on any one other community, library science, software engineering, et al. But I do think that if we did it well, it would be an opportunity to refine our skills at survey management. PierreGrenon: it will be tricky to define a specific domain and less appealing to a general audience. Amand's suggestion allows people from different domains to relate to the summit on the same level PierreGrenon: @amanda: I think the difficulty is primarily in diversity PeterYim: Dagobert: suggested "making sense out of data" as a topic AmandaVizedom: @Peter - I guess I'm thinking about each of the cases you've suggested, plus a few -- but I'm either up a level or too impatient. Instead of delving deep into one of those pairings, I'd like to look at the patterns. It would be natural for folks to do some deeper delving within the pairings that might run parallel and beyond. EricLindahl: (aside, a scientist I worked with responded to my suggestion of ontology in software with 'so you believe God is enumerable?') RexBrooks: @Eric: did you offer him an ontology of religiosity? RexBrooks: religiosity in software? EricLindahl: @Rex we're just sorting bits in a finite string, is his point I believe. PeterYim: @Amanda - not really in the case of "Toward Ontology-based Standards" (maybe we can't craft such a theme when putting other communities together with ontologists RexBrooks: @Eric: Ahh... AmandaVizedom: Perhaps one reason I think this is not harder is that I think the specific pairings may assume uniformity that isn't there. It seems very probable to me that when you dig into any one of these pairs, you find complexity that reflects the broader diversity. PierreGrenon: @amanda: yes, the risk is getting into fights, my proposal was too confrontational in that respect, obviously people will disagree. If we start from what people do and what are their requirements, we can perhaps have different answers of the practical utility of an ontological approach. EricLindahl: (everyone is shy all of a sudden) AmandaVizedom: Example: ontologies and libraries includes distinct activity- and interest- based groups around search and retreival, metadata and digital curation, text processing, ... more I think. EricLindahl: And maybe get sponsors RexBrooks: @Eric: I think none of us is fond of saying things twice? Once written, twice shy. PierreGrenon: @amanda: too abstract, can you explain 'pairs'or give an example? (i'll be reading upward meanwhile in case i missed smthg) PierreGrenon: oh nevemind EricLindahl: I second Amanda's real world complexity POV EricLindahl: General ontology design patterns for dealing with real world complexity in ontology development? PierreGrenon: @amanda: yes, "where can there be trade-offs" could be the theme EricLindahl: Ontology design heuristics EricLindahl: Does anyone use CLIPS or Flora-2 in ontology research? EricLindahl: But actually they have sub-minimal ontological intersection? RexBrooks: I think Amanda's points can be addressed in the context of Ontology Applications and Tools: (the domain considerations for best practices and methodologies can go in this explanatory passage.) JulitaBermejoAlonso: the ontology-based or ontology-driven label: what does it really mean? doug foxvog: The issue of people from different fields arguing, is often that they use the same terms in different ways. EricLindahl: Thank you all! AmandaVizedom: I will try to make a clearer proposal offline. doug foxvog: This does not mean that either is wrong, they are labeling different concepts with the same term. EricLindahl: Semiotics (just had to say it) PeterYim: Thanks everyone PeterYim: -- session ended: 2010.04.15-12:15pm PDT -- EricLindahl: So, that's the end of our Ontology party. Anyone need a ride home? PeterYim: One last time ... please endorse the Communique if you haven't already done so - http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2010_Communique and http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2010_Communique#nid2CUN