ppy/OntologySummit2014-s02_chat-transcript_unedited_20140123a.txt ------ Chat transcript from room: summit_20140123 2014-01-23 GMT-08:00 [PST] ------ [8:46] anonymous morphed into ShahrulAzmanNoah [9:02] PeterYim: . [9:03] PeterYim: Welcome to the = OntologySummit2014 session-02 Track-A: Common Reusable Semantic Content-I - Thu 2014-01-23 = Summit Theme: Summit Theme: OntologySummit2014: "Big Data and Semantic Web Meet Applied Ontology" Track-A Focus: Common Reusable Semantic Content Session Topic: Use and Reuse of Semantic Content - The Problems and Efforts to Address Them Session Co-chairs: * Mr. MikeBennett (EDM Council, Hypercube) * Ms. AndreaWesterinen (Nine Points Solutions) & * Dr. Gary Berg-Cross (SOCoP; Knowledge Strategies) Briefings: * Dr. GaryBergCross (SOCoP) - "Use and Reuse of Semantic Content: The Problems and Efforts to Address Them - An Introduction" * Professor PascalHitzler (Wright State U) - "Towards ontology patterns for ocean science repository integration" * Ms. AndreaWesterinen (Nine Points Solutions) - "Reuse of Content from ISO 15926 and FIBO" * Ms. MeganKatsumi & Professor MichaelGruninger (U of Toronto) - "Reasoning about Events on the Semantic Web" Logistics: * Refer to details on session page at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2014_01_23 * (if you haven't already done so) please click on "settings" (top center) and morph from "anonymous" to your RealName * Mute control (phone keypad): *7 to un-mute ... *6 to mute * Attn: Skype users ... see: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2014_01_23#nid44BM ** you may connect to (the skypeID) "joinconference" whether or not it indicates that it is online (i.e. even if it says it is "offline," you should still be able to connect to it.) ** if you are using skype and the connection to "joinconference" is not holding up, try using (your favorite POTS or VoIP line, etc.) either your phone, skype-out or google-voice and call the US dial-in number: +1 (206) 402-0100 ... when prompted enter Conference ID: 141184# ** 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: if the dialpad button is not shown in the call window you need to press the "d" hotkey to enable it . == Proceedings == . [9:25] anonymous morphed into JohnYanosy [9:27] anonymous morphed into Pascal Hitzler [9:28] LeoObrst: Folks, unfortunately I must leave after first hour. [9:28] PeterYim: go it, Leo ... glad you can still make a portion of this session [9:30] anonymous2 morphed into James Overton [9:31] anonymous2 morphed into AliHashemi [9:32] anonymous2 morphed into MariaPoveda [9:32] anonymous2 morphed into Conrad Beaulieu [9:33] anonymous2 morphed into JimSolderitsch [9:33] anonymous3 morphed into MikeCummens [9:33] James Overton morphed into JamesOverton [9:33] anonymous4 morphed into VeruskaZamborlini [9:33] Harold Boley morphed into HaroldBoley [9:34] anonymous2 morphed into JamesWilson [9:35] PeterYim: == MikeBennett starts the session on behalf of the Track-A co-champions ... see slides under: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2014_01_23#nid44B7 [9:36] anonymous2 morphed into MariaPoveda [9:37] anonymous2 morphed into RichardBeatch [9:38] anonymous3 morphed into JacobusGeluk [9:38] anonymous2 morphed into NaicongLi [9:41] PeterYim: == GaryBergCross presenting ... [9:41] Pascal Hitzler: While my talk will of course be self-contained, it is in a sense a different slice of the same storyline as presented at ontolog last November: http://ontolog.cim3.net/cgi-bin/wiki.pl?IAOA_SWAO_ConferenceCall_2013_11_25 [9:42] Pascal Hitzler: I also regret that I will have to leave shortly after my talk because of teaching duties. [9:44] anonymous2 morphed into Ramin Ayanzadeh [9:47] anonymous2 morphed into JulienCorman [9:51] anonymous2 morphed into Max Gillmore [9:51] MikeBennett: @Peter, the VNC hopper seems to be showing the slide not full screen put placed on top of your overall screen (bit of Session Page showing below) [9:54] TaraAthan: slide 9 - an opportunity for alignment with a quantities/units of measure ontology? [9:56] PeterYim: == PascalHitzler presenting ... [9:56] AndreaWesterinen: @TaraAthan Yes, I touch on this a bit also in my presentation. [9:56] QuentinReul: @GaryBergCross: Do you mean that SKOS doesn't encode the formal semantics of similarity? [9:57] PeterYim: @MikeBennett ... I cannot reproduce your problem ... anyone else has that problem Mike is mentioning? [9:57] AliHashemi: Hi Gary, I find question 7 in your supplementary material to be very interesting - "Is reuse about semantics alone or should it also address reasoning and data analytics?". Could you elaborate a bit the difference between how one accesses the semantics of a concept vs reasoning about a concept? [9:57] MikeBennett: @Peter it's OK now. [9:58] Gary Berg-Cross: @Quentin I'm not sure that SKOS covers similarity is as useful a way as needed. It is a thesaurus level and doesn't get in how things are similar. [9:59] Gary Berg-Cross: Ali - Mike Gruninger's talk should provide some context for this issue. So let's hear that first. [9:59] QuentinReul: @GaryBergCross I agree, but it was not intended to provide simialrity at the conceptual level [9:59] PeterYim: @MikeBennett - it's probably because GaryBergCross' slide format (aspect ratio) does not coincide with the screen's ... I show a huge "black stripe" at the top and the bottom, but I suppose my "black" could have been "pixels from the previous image" in your case [9:59] MikeBennett: NSF = National Science Foundation (in the US), for those who don't know it. [10:00] MikeBennett: @Peter that will have been it- the slide went full width but we saw top and bottom of your screen. Or I did anyway. [10:00] QuentinReul: @GaryBergCross some have argued that thesauri are sufficient for most business needs, which I totally disagree with [10:01] anonymous3 morphed into JuanGomezRomero [10:03] Gary Berg-Cross: @Quentin Yes, I agree with you, I think that it is part of the need for deeper knowledge on concepts P and S. [10:03] AliHashemi: @Pascal, don't both extremes have their difficulties in their reuse? [10:03] AliHashemi: A theory that is underspecified may have a lot of implicit assumptions in how the model is used [10:03] AliHashemi: Which would lead to unintended errors in re-use as well, no/ [10:04] QuentinReul: @GaryBergCross Did some work for my PhD where I relied on DL axioms to analyse similarity between concepts, and then re-used construct to represent the similarity. Have you heard of EDOAL (http://alignapi.gforge.inria.fr/edoal.html)? [10:05] Gary Berg-Cross: @Ali I hope that Pascal has time to respond to this, he has to go to another meeting, but perhaps Jano who is also on the call would offer some ideas here. [10:05] Gary Berg-Cross: @Quentin No that is a new one to me. Thanks.... [10:05] MikeBennett: This (Hitzler, slide 9) ODP of Event is identical to the one in FIBO. [10:06] Gary Berg-Cross: @Mike Ah! A building block for reuse already! [10:06] MikeBennett: (except we refer to the concepts of place and time, rather than strings) [10:08] anonymous3 morphed into MatthewLange [10:08] MikeBennett: Oh, and Place need not be spatial - there are events which occur at a virtual place, e.g. a security is issued on the Global Bonds Market. [10:08] SimonSpero: It's not just a Logical Form of Action Statements; it's THE Logical Form of Action Stations [10:09] SimonSpero: (Captain Donald Tiberius Davidson) [10:09] JohnMcClure: is a template, not a pattern. A pattern is more about the way multiple things are put together. [10:10] anonymous1 morphed into GenZou [10:10] SimonSpero: Davidson paper (scanned from "Action and Events", I assume) http://isites.harvard.edu/fs/docs/icb.topic638346.files/Davidson1967.pdf [10:12] AmandaVizedom: @Pascal, that's a very good and important point: patterns and modularity are related. patterns increase the degree to which chunks of modeling can be plugged into each other with predictable results. That enables modularity and increases reusability. [10:13] MatthewLange: These terms, "patterns", "model", "predictability", "axioms" etc, should be captured in an ontology, if they are not already (perhaps I don't know about the source??) [10:14] MikeBennett: @Matthew Good agenda for a meta-ontology. [10:14] SimonSpero: @Pascal :Can either be expressed in OWL ? [10:15] MichaelGruninger1: @Gary Berg-Cross: do you see a distinction between an ODP and a generic ontology? Hitzler slide 10 is part of an event ontology. You could call it a pattern insofar as it is a pair of axioms that are common across many of the existing event ontologies [10:15] MatthewLange: I can't see live slides, is VNC server down? [10:15] MikeBennett: VNC is OK here. [10:16] MichaelGruninger morphed into MeganKatsumi [10:17] Gary Berg-Cross: @MichaelGruninger1 I think that ODPs are often a PART of a top=levle or domain ontology, but selectively and modualrly so. So many patterns are pieces of DOLCE [10:17] SimonSpero: @Pascal specifically: aren't Property Chains constrained to Object Properties? [10:17] MatthewLange: @MikeBennett indeed, does such an animal exist? of course, it should also contain "description logic" "OWL" , etc, etc [10:17] PeterYim: @MatthewLange & All, vnc is strictly optional, and people behind corporate firewalls would run into problems unless they try to get ready for it beforehand ... just run the slides on your own desktop - the speakers will prompt everyone to advance slides [10:19] QuentinReul: @PascalHitzler Are these patterns added to the ODP repository (http://ontologydesignpatterns.org/wiki/Main_Page)? [10:21] MichaelGruninger1: @Pascal Hitzler: Do you see an ontology pattern as distinct from a subtheory of an ontology (along the lines of Gary's comment)? [10:21] PeterYim: ^ http://ontologydesignpatterns.org/wiki/Main_Page [10:21] PeterYim: @ALL ... when posting a link here, please leave a blank space before and after the url (the chat-room tends to include your commas or close parentheses as part of the url! [10:22] SteveRay: +1 on QuentinRuel's question [10:22] ScottHills: @PascalHiztler: Regarding question on slide 17: If you generalize the concepts of TemporalThing (time) and SpatialThing (place) to allow ranges (time interval, spatial extent), I don't think there's any problem considering both applicable to a cruise. [10:22] Pascal Hitzler: I have a few minutes, will try to catch up with chat discussion a bit [10:22] PeterYim: == AndreaWesterinen presenting ... [10:23] JohnYanosy: Sets of patterns to create views is an interesting approach to a situation where a massive ontology might not be feasible to understand. [10:24] MatthewWest: @Ali: I think there is a difference between being able to reuse the pattern, and being able to compare and use together the results of the reuse of the ontology. Reuse of the ontology is eased by fewer axioms, but reuse and comparison/integration of the results of using the ontology is improved if you have more axioms (less chance for unintended models). So it depends what you are trying to achieve. [10:25] Pascal Hitzler: @Ali: yes, both extremes have problems regarding reuse. Implicit assumptions are a problem of course, yes. I would probably say that implicit assumptions are in fact ontological commitments which have not been communicated. Of course that causes problems. [10:25] AliHashemi: @Matthew, I agree. [10:26] Gary Berg-Cross: @JohnY Yes, it has bee widely noted that domain people (and perhaps even we knowledge engineers) have difficulty going through and understanding a large ontology and then are faced with how to pick out useful pieces and exclude pieces that might have axioms they don't agree with. A KE can't do it w/o domain knowledge and a domain person cant't understand the formalisms. [10:26] QuentinReul: @MatthewWest Is it not adding fewer constraining axioms that would reduce complexity and thus increase re-use? [10:26] AliHashemi: It seems that at the extreme, a highly modularized ontology might blend into an ontology pattern? [10:26] Pascal Hitzler: @Mike: I wasn't aware that this (slide 9) is as in FIBO. Thanks for the pointer :) [10:27] Gary Berg-Cross: @Ali The modules in such an ontology might. [10:27] AliHashemi: @Pascal, also agree. @Quentin, it depends on how the term is being used. If the usage of the term belies unstated assumptions, then it is a superficial type of "easier reuse" [10:27] Pascal Hitzler: @John: I'm not sure about the exact difference between template and pattern, if any. [10:27] MikeBennett: @Gary the domain people can understand the formalisms if they are presented and explained in terms of set theory. Where that falls down is with OWL Restrictions, which require some notational footwork to re-frame them as refinements of re-uses of a property. [10:28] PeterYim: @AndreaWesterinen ... [ref. question about previous coverage of iRing in this community] MatthewWest gave a talk about it at the OntologySummit2009_Symposium - see: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2009_Symposium#nid1VZT [10:28] anonymous2 morphed into JennSleeman [10:28] Pascal Hitzler: @Amanda: thanks [10:28] Gary Berg-Cross: A MikeB Good lesson from your work out there with people... [10:29] LeoObrst: @Ali: yes, I think ontology modules coming mostly top-down meet at ontology patterns coming mostly bottom-up. I think in fact that eventually those patterns, after greater refinement, may correspond to those ontology modules. [10:29] Pascal Hitzler: @Matthew: I agree, although it may be too early to try capture these in an ontology. We definitely need to get a good understanding of notions related to pattern, and of possible relationships between patterns. [10:29] Pascal Hitzler: @Simon: Sorry missing reference. re. expressivity in OWL - which slide? [10:30] QuentinReul: @MikeBennett Did you use graphical notations to represent the info to Subject-Matter Experts (SMEs)? If so, what notation did you use? I have found that set theory was sometimes already too difficult for some SMEs [10:30] MatthewLange: @Everyone, could you please use full names, there are two Matthews here, it's getting confusing ;-) [10:30] SimonSpero: @pascal: Slide 14 [10:31] Gary Berg-Cross: Semantic overlap seems to be another way of describing the concept similarity issue. [10:31] Pascal Hitzler: @MichaelGrueninger: Regarding ODP versus ontology: I would argue that we need to make sure that the patterns are exposed. If you publish only a large ontology, but have lost how it emerges from patterns, then reuse is imo made much more difficult. [10:31] MatthewWest: @Pascal: When we were developing ISO 15926, one of our primary concerns was to limit the options for how things could be modelled, but at the same time to allow anything to be modelled - the aim being reuse of data. These two desiderata drove a lot of the choices we made. [10:31] ElieAbiLahoud: +1 @LeoObrst reply to Ali [10:32] MikeBennett: @Quentin yes we did - but for this we used a cut-down implementation of OWL which did not include restrictions. We created a nearly Visio-like presentation of nodes and edges (for classes and properties), with datatype properties as textual entries within a box (very UML-like but without any UML notation). Now exploring how to render Restrictions in a similar way. [10:32] PeterYim: further to MatthewLange's [10:30], please cite full (WikiWord name), every instance that you abbreviate mean another burden on whoever is going to spend time cleaning up the chat-transcript for the archives ... Thanks in advance! [10:32] Pascal Hitzler: simon: I think you're right that property chains are constrained to object properties (my bad). However, that may just be OWL (not sure there's a deeper reason to not allow a datatype property at the end of a chain - it was probably just left out of the standard). [10:32] SimonSpero: @pascal: has an Object Property as a subproperty of a Data PRoperty [10:33] PeterYim: ^^^ @All [10:33] SimonSpero: @Pascal: There were some concerns about... decidability [10:33] Pascal Hitzler: our patterns are still work in progress, so they are not yet published anywhere. [10:33] MikeBennett: @Andrea when you were finding legal ontologies, did you find any ontologies dealing with contracts? [10:33] QuentinReul: @MikeBennett I have tried a very similar approach and it didn't quite work. The main issue is trying to express that certain things are together. Could you share some of your diagrams? [10:33] MichaelGruninger1: @Pascal Hitzler; Perhaps people should stop publishing LARGE ontologies, and focus on publishing small ontologies that get combined as modules of larger ontologies. COLORE is full of small modules which get combined into larger ontologies. In particular, you can find a modularization of DOLCE [10:33] Pascal Hitzler: (that was @ScottHills)) we should soon be there though [10:33] Pascal Hitzler: @JohnYanosy: yes! [10:34] Pascal Hitzler: @MatthewWest: what you say is part of a discussion which we really need to have. We do not yet understand these trade-offs well enough yet. [10:35] QuentinReul: @MikeBennett I would especially be interested in your current thinking about representation of restrictions [10:36] MikeBennett: @Quentin the diagrams are at www.hypercube.co.uk/edmcouncil - the current diagrams are slightly different as we have made minor changes to the metamodel (so disjoints and inverses are now dashed not solid lines). These diagrams required a 1 hour explanation to business domain experts before they could participate. [10:36] MikeBennett: @Quentin I can mail you my current proposals (which the more technical people don't like!) [10:37] JohnMcClure: Slide 13: Too much complexity, mindboggling tens of thousands of classes... does she mean tens of thousands of Properties? [10:37] Pascal Hitzler: @SimonSpero: regarding slide 14: the lower of the two axioms is not expressible in OWL. It has been looked at, though, in research around description logics. IIRC, even for relatively small logics it causes undecidability if used in an unrestricted way. [10:37] ElieAbiLahoud: @MichaelGruninger1 "Perhaps people..." YES, like regrouping code in functions, modules, etc... [10:38] PeterYim: ^ http://www.hypercube.co.uk/edmcouncil ... (ALL: kindly use fully qualified url ... ) [10:38] QuentinReul: Did anyone look at www.bit.ly/OpenContractingData ? It seems to be a new standard to represent contracting info [10:38] MikeBennett: Noted - sorry! [10:38] Pascal Hitzler: @SimonSpero: thanks for the remark about decidability, I'd have to look into this. Very interesting in fact. [10:38] MikeBennett: http://www.bit.ly/OpenContractingData [10:39] TaraAthan: Regarding the suggestion of small modules that are reused - often axioms use terms that are defined in other modules. When importation of these dependencies is required, you get modules that are not so small anymore. When importation of such dependencies is optional, you may get a shift in semantics. And then there is the question of circular dependencies... [10:39] SimonSpero: @Pascal: Right - but there were concerns for some reason about data properties... I need to try and find the email thread (obviously a DPE can only be the terminal property in a chain) - the discussion seemed to peter out [10:39] QuentinReul: @MikeBennett I would like that (if not propriatory). I may not like it, but I would not be the main consumer of it [10:39] Pascal Hitzler: @MichaelGruninger: small versus large ontologies: Perhaps, if publishing large ontologies, make sure that the patterns used are still well exposed and easily located. [10:39] PeterYim: @MikeBennett:, No apologies necessary ... (just trying best to cope with idiosyncrasies of the tools we are living with, and still be lazy :) [10:40] Pascal Hitzler: @SimonSpero: if you find a link, that would be tremendous! [10:41] Pascal Hitzler: I'm afraid I have to leave now for teaching. Thanks a lot to all, and the discussion. Feel free to email me if you have further toughts [10:41] AliHashemi: @Tara, that raises the questions of whether patterns can be that self-contained... And it gets back to the crux of how modules or theories can extend one another [10:41] PeterYim: Bye, Pascal ... thank you for the great talk! [10:41] SimonSpero: @Pascal @Tara: Quasi second order axioms A la CL/IKL/CycL can capture a lot. CycL macros are also nice [10:41] MikeBennett: Thanks Pascal - and thanks for following up the questions. [10:41] MichaelGruninger1: @Tara: Agreed. Ontology designers and users need to be aware when another ontology is being conservatively extended and when it is a nonconservative extension. [10:43] MatthewLange: @MichaelGruninger1, I agree with you about small ontologies, except that sometimes we need to develop knowledge models around the larger value propositions possibly gained by modeling contained domains [10:43] JohnMcClure: she means annotation properties are separate from the text/object properties [10:43] SimonSpero: @MikeBennet: Agreement Technologies (in Springer Law, Governance and Technology series): http://www.springer.com/computer/ai/book/978-94-007-5582-6 [10:45] AnatolyLevenchuk: ISO 15926 self-education sequence -- http://levenchuk.com/2012/10/01/iso-15926-self-education-sequence/ [10:46] anonymous2 morphed into CyndyChandler [10:46] PeterYim: == MeganKatsumi-MichaelGruninger presenting ... [10:47] MatthewWest: @Andrea: I would be very surprised if anyone reused all of ISO 15926. Typically, projects use an extended subset, some of what they want is there, and some is not there. The idea is to develop the extensions with reuse in mind, so they can be added back to the whole. ISO 15926 is intended to be reusable and extensible. [10:51] AndreaWesterinen: @JohnMcClure -Slide 13 - I was talking about the full process industry ISO - which makes many subclasses. Sorry for not being clear. [10:51] SimonSpero: Ryan Shaw's Dissertation (LODE) is online at http://aeshin.org/dissertation/ [10:51] SteveRay: @Andrea: Where is the best place to find a copy of psl.owl? [10:52] MichaelGruninger1: @SteveRay: we will be uploading all of the OWL axiomatizations into the OntoHub repository [10:52] AliHashemi: @Megan and Michael, are these extensions available online? [10:52] anonymous2 morphed into BartGajderowicz [10:53] AmandaVizedom: @MeganKatsumi: re: slide 10, can you say something about why you chose SWRL for this? [10:53] VictorAgroskin: @Andrea: Looks like you have used mostly data model (ISO 15926-2), not Reference Data Libraries developed for it? [10:54] FrankLoebe: @Megan/Michael: Are the translations from first-order PSL to OWL and SWRL (partially?) automatic? [10:54] AndreaWesterinen: @Victor We looked at both, but yes, mostly the data models. [10:54] MichaelGruninger1: @Amanda: we wanted a language that combined both OWL and rules [10:55] AndreaWesterinen: @Amanda We also are using SWRL. We have tools like Stardog that input and reason with SWRL. [10:55] MichaelGruninger1: @FrankLoebe: the translation definitions between the ontologies are manually generated. [10:56] AndreaWesterinen: @MikeBennett We looked at the Public Contract Ontology,https://code.google.com/p/public-contracts-ontology/ [10:57] AndreaWesterinen: @MikeBennett ... and ontologies/concepts summarized in the IOS Press book, "Law, Ontologies and the Semantic Web" [10:58] AndreaWesterinen: @MikeBennett For example, they also discuss design patterns in Chapter 3. [10:58] AmandaVizedom: @MichaelGruninger, @MeganKatsumi & AndreaWesternen: One reason I asked is that it's counter to what I perceive as dominant approach in semweb. Seems as if many OWL users bypass SWRL and use SPARQL queries, and scripts executing them, in place of declarative rules. This has some obvious shortcomings (loses benefits of being declarative, for example), but is common. SWRL approach caught my attention. [10:58] MikeBennett: @Andrea - cool! Thanks. [10:58] PeterYim: @MeganKatsumi & MichaelGruninger, is there documentation that elaborates on the results (at least to the extent supporting the "preliminary results" cited on your slide#15? [10:58] MatthewWest: @Andrea and Victor: Interestingly it is the data model that contains the key ontology patterns and commitments, and the Reference Data Library that contains the classes that bring the specificity to the patterns. So this is not unexpected when used outside the Process Industries. [10:59] AndreaWesterinen: @SteveRay, @MeganKatsumi I think that Megan has to answer the psl.owl question. [11:00] AndreaWesterinen: @MatthewWest That was my experience, so thanks for confirming it! [11:01] PeterYim: @MeganKatsumi & MichaelGruninger, great talk! ^ [11:02] SimonSpero: @MichaeGruninger, @Megan Ryan Shaw asked whether you also looked at F ( http://west.uni-koblenz.de/Research/ontologies/events/event-model-f-kcap.pdf ) [11:02] AndreaWesterinen: Previous discussion items about "small" patterns within larger ontologies/models seems like a good way to go. AND ... including mappings where semantics are taken, mapped and extended. [11:02] PeterYim: == Q & A and Open Discussion ... [11:03] SimonSpero: @MichaeGruninger, @Megan: And points out that LODE was not explicitly designed for reasoning, so the evaluation is meaninful [11:03] Dennis Wisnosky: Time at the end is a shock. [11:03] MeganKatsumi: @Peter Yim: We're finishing up a draft to submit to the Journal of Web Semantics [11:04] MeganKatsumi: We can also distribute some preliminary notes [11:05] JohnMcClure: @andrea - terrific talk thank you for the time you invested! I was noting that Slide 11 shows three lists of properties, not classes [11:07] PeterYim: @MeganKatsumi, thank you ... please contribute that in time because your research (and the results, even if preliminary) is so relevant to this Summit (the journal paper timeline may work off a different pace) [11:07] MatthewWest: @Andrea: if you want to work together on elucidating events/activity in ISO 15926 during the summit, I would be happy to help with that. [11:07] Gary Berg-Cross: PASCAL showed building the cruise pattern using the trajectory pattern which in turn used an event pattern. [11:08] MichaelGruninger1: @SimonSpero: Yes, we had planned on including the F-model of events; one of the obstacles was its tight integration with DOLCE UltraLight. This hearkens back to the earlier comments about reusing modules of ontologies rather than the entire ontology [11:08] AndreaWesterinen: @JohnMcClure Yes, slide 11 is from FIBO. Their object/data/annotation properties were very well thought out and defined. [11:09] AndreaWesterinen: @MatthewWest That would be great!! Let's talk off line and then put some thoughts out on the mail list. [11:09] MichaelGruninger1: @MikeBennett: I missed what you just said about activities and events -- could you repeat in the chat? [11:10] VictorAgroskin: @Matthew, Andrea: I've noticed the wish to extend ISO 15926 ontology. It will be beneficial if Andrea extends reference data as it is prescribed by ISO 15926, not data model. Then we (ISO 15926 community) will be able in turn to reuse content developed by another community. [11:11] MikeBennett: We considered Event to be something with a time and a place (and, implicitly some cause therefore some causal agent). Extending this to an Event in which there is some Actor - what that equates to is an Act or Activity, and Activities becoem (along with Events) the building blocks of a Process. And so on. At this point the pattern diverges from ones that others have usd, but it seems to match the common sense meanings. [11:12] MikeBennett: @Michael also these concepts have mappings to concepts in REA under different names (e.g. REA "Event" is an Activity). [11:12] AndreaWesterinen: @Victor We are looking to extend similar to Part 4, which is the reference data. No? [11:13] JohnYanosy: A question for Andrea regarding Network management ontology - in this context there are network elements with static performance capabilities and instances in time of dynamic variables, which affect or influence higher level network performance variables, such as cost, availability, utilization, etc. Many of these higher level concepts are not logical results of elements but rather require higher level mathematical processing to relate them. What role do you see for an ontology in this context? Possible external mathematical reasoning models may be required to create network level performance instances related to dynamic network element instances? [11:13] MatthewWest: @Victor, Andrea: I agree. There is supposed to be enough data model, though there is the possibility of finding other patterns in class_of_relationship which you might want to give more prominence to. So a way to discover and document those for reuse might be useful. Of course the templates are supposed to cover that. [11:13] AndreaWesterinen: @Victor and @MatthewWest But, there may be data model extensions. [11:14] Gary Berg-Cross: `I agree with what Mike G just said about the limitations of building blocks. It remains a challenge to understand the limits of reuse and under what conditions these are reasonable. [11:15] PeterYim: @Gary - if you can do a local mute, then I will not have to mute your line (we are hearing the "thump" when you type, unless you are muted) [11:15] MatthewWest: @Andrea: extensions to the data model are not in possible, but there have been no updates in 10+years now. [11:15] AliHashemi: @Simon, but all. Since you mentioned in your comment that the LODE was designed without reasoning in mind, could elaborate how one accesses the semantics of a concept or a term and how that is different from reasoning about the term/concept? [11:16] Gary Berg-Cross: Peter OK... [11:16] PeterYim: @Gary ... your line is not muted from the bridge control now [11:17] SimonSpero: AliHashemi: I believe the focus was on Intentional descriptions - see Ryan's dissertation above [11:17] AndreaWesterinen: @JohnYanosy Do you want to collaborate a bit offline and discuss this further? [11:19] AmandaVizedom: @MichaelGruninger: some of the philosophers are pragmatists. ;-) [11:20] MatthewWest: @Amanda: Yes, I sometimes describe myself as an applied philosopher :-) [11:22] ToddSchneider: To clarify, the ability to reuse an ontology is dependent on the 'congruence' of the set of intended interpretations. And (hopefully) any inferencing will preserve these interpretations. [11:22] PeterYim: MichaelGruninger: would be good to address the questions (crafted mainly by AndreaWesterinen) on slide#15 of GaryBergCross' slide deck [11:22] MichaelGruninger1: @AndreaWesterinen, MikeBennett, Gary Berg-Cross -- the track questions on slide 15 of Gary's slide are great! [11:23] PeterYim: GaryBergCross: soliciting from the community - what do you want to hear on our next session of this Track-A: Common Reusable Semantic Content [11:23] AndreaWesterinen: @Peter I can start a thread on each question. [11:23] VictorAgroskin: @Andrea: ISO 15926 data model has some limitations because of ontology foundations. One obstacle you can meet in network-related modelling can be in the modelling of "information" domain. ISO 15926 just is not designed to transparently type things like "byte", "package", "file", not to say "protocol" or "handshake". If there is some established ontology for that - I'll really recommend to use it, just making a link to ISO 15926 ontology at very high level. [11:24] AmandaVizedom: @MatthewWest: I sometimes describe myself as an applied epistemologist, entirely straightforwardly, but of course that phrase is meaningful to even fewer people than "ontologist" is. :-) [11:24] AndreaWesterinen: @Victor I agree, but the team was looking to reuse tools that were designed for the ISO models. We can start from class_of_class :-). [11:24] MikeBennett: Please suggest speakers for the second Track A session. [11:25] JohnYanosy: Andrea I would love to collaborate off-line. My email is jyanosyjr [at] gmail.com [11:25] MikeBennett: Modularity and re-usability as an issue to explore (see Q4 on Slide 15 of Gary's slides) [11:25] PeterYim: @AndreaWesterinen ... by all means, that's what the [ontology-summit] discussion list and the wiki page at http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2014_Common_Reusable_Semantic_Content_CommunityInput are for [11:25] AndreaWesterinen: @Victor Also byte, packet, etc. are more units than specific classes. [11:26] SteveRay: So I'm gathering that the main places to find collections of reusable ontologies or patterns are: http://ontologydesignpatterns.org/wiki/Ontology_Design_Patterns_._org_%28ODP%29, http://www.ontohub.org/, and http://ontolog.cim3.net/cgi-bin/wiki.pl?OpenOntologyRepository. Yes? Other places to start? [11:26] Gary Berg-Cross: MikeG I like this idea of exploring modularity... BTW Aldo gangemi will be a future speaker in Track C Bottlenecks, I believe.. [11:26] AndreaWesterinen: @MikeB and MichaelG +1 on modularity [11:27] MichaelGruninger1: @SteveRay: Oliver Kutz is coordinating efforts in Track E on uploading specific ontologies that arise thorughout the Summit into the open ontology repositories, particularly OntoHub [11:28] MichaelGruninger1: oops -- I meant Track G: Community Resources [11:28] MariaPoveda1: I guess you already know about http://lov.okfn.org for looking for ontologies [11:28] PeterYim: @SteveRay - try also: http://oor.net ... note also that an ontology repository is being stood up to collect ontologies that crosses path with this Summit - see: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2014#nid442O [11:29] JohnYanosy: I think there are some lessons to be learned from the Object Oriented community who had similar problems of scope of classes by developers. In early years many very large objects were created. Are there similarities? [11:29] Gary Berg-Cross: @JOHnM So modularity might be necessary but not sufficient. It may get us to a research space that we can understand better getting there is steps. [11:31] MatthewWest: @Andrea: the problem with modularity is that, as you found with FIBO, you find that if you try to use one bit, you quickly find you need to use many other modules to do anything useful. Also the separate modules have to be consistent with each other. So I think it is more useful to think in terms of core ontologies, or foundation ontologies, and domain ontologies that are essentially class libraries that give you the specific vocabulary for a domain. [11:32] Gary Berg-Cross: I have to get off for another meeting.. Thanks everyone. Let's keep the discussion going on the forum. [11:33] SimonSpero: @andreaweterinen: So are you going to develop a BER/PER serialization for RDF? [11:33] ToddSchneider: AndreaWesterinen, do you have an architecture for network management ontologies? [11:33] AmandaVizedom: This is the direction addressed by some repositories, but only currently works within those repositories. LOV most oriented toward reuse, perhaps. [11:33] SteveRay: For the record: I was asking for a "WordNet for ontologies" [11:34] MeganKatsumi: @SteveRay: We also have some experiences in looking for ontologies that capture the semantics of actors -- we can continue this discussion on the Track list [11:34] PeterYim: @ALL: if you are not subscribed to the [ontology-summit] mailing list yet, please do so - http://ontolog.cim3.net/mailman/listinfo/ontology-summit (or drop me a line - peter.yim [at] cim3.com) [11:34] PeterYim: @all ... Session-03 will be up next Thursday- Thu 2014.01.30 (same time) - see developing details at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2014_01_30 [11:34] PeterYim: Reminder to those in the organizing committee, our 5th meeting coming up tomorrow - Fri 2014.01.24 - see: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2014/GettingOrganized#nid44G0 [11:35] AndreaWesterinen: We used annotation properties in our class definitions, wordNetSynonym, wordNetHypernym and wordNetHyponym, to capture semantics for our classes in the hopes of using this for mapping someday. [11:35] PeterYim: We still need volunteers as co-champions in Public Relations and Program Management, if you are interested, if indicate here, email offline, and come to our organizing committee meeting tomorrow [11:35] PeterYim: great session ... bye, everyone! [11:36] MikeBennett: Thanks all for a great session and some really directions for further exploration in the area of semantics re-use! [11:36] AndreaWesterinen: I will try to publish more about the network management work over the next few weeks. [11:38] JohnYanosy: Thanks for wonderful session. I will collaborate on some Network Management ontology issues on the listserv [11:36] PeterYim: -- session ended: 11:33am PST -- [11:36] List of attendees: AleksandraSojic, AliHashemi, AmandaVizedom, AnatolyLevenchuk, AndreaWesterinen, AndreaWesterinen1, Anne Thessen, BartGajderowicz, Bob Schloss, ChristineKapp, Conrad Beaulieu, CyndyChandler, DaliaVaranka, DanielMcShan, Dennis Wisnosky, DennisPierson, EarlGlynn, ElieAbiLahoud, FabianNeuhaus, Fran Lightsom, FrankLoebe, Gary Berg-Cross, GenZou, Harold Boley, HaroldBoley, HensonGraves, JacobusGeluk, James Overton, JamesOverton, JamesWilson, JennSleeman, JimSolderitsch, JoelBender, JohnMcClure, JohnYanosy, JuanGomezRomero, JulienCorman, KenBaclawski, Krzysztof Janowicz, LeoObrst, LianaKiff, MariaPoveda, MariaPoveda1, MatthewLange, MatthewWest, Max Gillmore, MeganKatsumi, MichaelGruninger, MichaelGruninger1, MikeBennett, MikeCummens, NaicongLi, NaicongLi1, NancyWiegand, OliverKutz, Pascal Hitzler, PeterYim, QuentinReul, QuentinReul1, Ramin Ayanzadeh, Ramin Ayanzadeh1, RexBrooks, Richard Martin, RichardBeatch, ScottHills, ShahrulAzmanNoah, SimonSpero, SteveRay, SteveRay1, Sunday Ojo, Sunday Ojo1, Sunday Ojo2, TaraAthan, TerryLongstreth, TimFinin, ToddSchneider, VeruskaZamborlini, VictorAgroskin, anonymous, anonymous1, anonymous2, anonymous3, anonymous4, vnc2 ------