From golux at comcast.net Fri Apr 2 04:59:53 2004 From: golux at comcast.net (Stephen Waterbury) Date: Fri Apr 2 04:36:51 2004 Subject: [step-os] Testing! Message-ID: <406D3999.2040302@comcast.net> Hi David and Lothar, Did you notice any weirdness when subscribing? I think the configuration wasn't quite right when I sent out the invitations, but it should be okay now. (Guess I should have tested a little more ... :) Cheers, Steve From david.price at eurostep.com Fri Apr 2 05:06:15 2004 From: david.price at eurostep.com (David Price) Date: Fri Apr 2 05:10:27 2004 Subject: [step-os] Testing! In-Reply-To: <406D3999.2040302@comcast.net> Message-ID: <000701c4189a$228fdbe0$2101a8c0@esukpc20> I received this test. The subscription email subject was weird though: confirm 040d1771697f46f41eabe5cdd2dda3742c72fa74 DP > -----Original Message----- > From: step-os-bounces@ned.gsfc.nasa.gov [mailto:step-os- > bounces@ned.gsfc.nasa.gov] On Behalf Of Stephen Waterbury > Sent: 02 April 2004 11:00 > To: step-os@ned.gsfc.nasa.gov > Subject: [step-os] Testing! > > Hi David and Lothar, > > Did you notice any weirdness when subscribing? > I think the configuration wasn't quite right when > I sent out the invitations, but it should be okay > now. (Guess I should have tested a little more ... :) > > Cheers, > Steve > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://ned.gsfc.nasa.gov/mailman/listinfo/step-os From golux at comcast.net Fri Apr 2 05:37:50 2004 From: golux at comcast.net (Stephen Waterbury) Date: Fri Apr 2 05:14:45 2004 Subject: [step-os] Testing! In-Reply-To: <000701c4189a$228fdbe0$2101a8c0@esukpc20> References: <000701c4189a$228fdbe0$2101a8c0@esukpc20> Message-ID: <406D427E.4070005@comcast.net> David Price wrote: > I received this test. > > The subscription email subject was weird though: > > confirm 040d1771697f46f41eabe5cdd2dda3742c72fa74 That's normal for a Mailman list. That's so you can just send a blank reply from most clients because they'll leave the cookie in the subject line. From lothar.klein at lksoft.com Fri Apr 2 06:21:17 2004 From: lothar.klein at lksoft.com (Lothar Klein) Date: Fri Apr 2 05:54:25 2004 Subject: [step-os] Testing! In-Reply-To: <406D3999.2040302@comcast.net> References: <406D3999.2040302@comcast.net> Message-ID: <87152770422.20040402132117@lksoft.com> Steve, For me all went well - I had no problems Lothar -- // Lothar Klein, LKSoftWare GmbH // Steinweg 1, 36093 Kuenzell, Germany // +49 661 933933-0, Fax: -2 // mailto:lothar.klein@lksoft.com http://www.lksoft.com Friday, April 2, 2004, 11:59:53 AM, you wrote: > Hi David and Lothar, > Did you notice any weirdness when subscribing? > I think the configuration wasn't quite right when > I sent out the invitations, but it should be okay > now. (Guess I should have tested a little more ... :) > Cheers, > Steve > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://ned.gsfc.nasa.gov/mailman/listinfo/step-os From david.price at eurostep.com Fri Apr 2 08:26:47 2004 From: david.price at eurostep.com (David Price) Date: Fri Apr 2 07:59:54 2004 Subject: [step-os] Testing! In-Reply-To: <87152770422.20040402132117@lksoft.com> Message-ID: <001601c418b6$262044e0$2101a8c0@esukpc20> Steve, So, is this ready for me to tell my colleagues about? OR... should I await a broadcast on the usual exploders before telling them? Cheers (and again.. THANKS!), David > -----Original Message----- > From: step-os-bounces@ned.gsfc.nasa.gov [mailto:step-os- > bounces@ned.gsfc.nasa.gov] On Behalf Of Lothar Klein > Sent: 02 April 2004 12:21 > To: Stephen Waterbury > Cc: step-os@ned.gsfc.nasa.gov > Subject: Re: [step-os] Testing! > > Steve, > > For me all went well - I had no problems > > Lothar > > > -- > // Lothar Klein, LKSoftWare GmbH > // Steinweg 1, 36093 Kuenzell, Germany > // +49 661 933933-0, Fax: -2 > // mailto:lothar.klein@lksoft.com http://www.lksoft.com > > Friday, April 2, 2004, 11:59:53 AM, you wrote: > > Hi David and Lothar, > > > Did you notice any weirdness when subscribing? > > I think the configuration wasn't quite right when > > I sent out the invitations, but it should be okay > > now. (Guess I should have tested a little more ... :) > > > Cheers, > > Steve > > _______________________________________________ > > step-os mailing list > > step-os@step.nasa.gov > > http://ned.gsfc.nasa.gov/mailman/listinfo/step-os > > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://ned.gsfc.nasa.gov/mailman/listinfo/step-os From golux at comcast.net Sun Apr 4 15:15:06 2004 From: golux at comcast.net (Stephen Waterbury) Date: Sun Apr 4 14:51:52 2004 Subject: [step-os] Testing! In-Reply-To: <406DA500.5040304@comcast.net> References: <001601c418b6$262044e0$2101a8c0@esukpc20> <406DA500.5040304@comcast.net> Message-ID: <40705EBA.3050008@comcast.net> David Price wrote: > So, is this ready for me to tell my colleagues about? OR... should I await a > broadcast on the usual exploders before telling them? Absolutely! Go ahead and tell your colleagues about it. I will announce it on the PDES, Inc. lists, but I wasn't planning to announce it elsewhere. However, you should feel free to announce it on other lists that you think should know about it. Cheers, Steve From golux at comcast.net Mon Apr 5 18:55:36 2004 From: golux at comcast.net (Stephen C. Waterbury) Date: Mon Apr 5 18:28:13 2004 Subject: [step-os] And the kick-off: STEP and ontologies Message-ID: <4071E3E8.3@comcast.net> Greetings, fellow STEP-OSers! Judging by our membership (which you can see if you go to http://step.nasa.gov/mailman/listinfo/step-os and enter your address and password), I think we have a quorum. Also, all messages to this list go into our archive: http://www.step.nasa.gov/pipermail/step-os ... so late-comers will have access to our discussion history, anyway. ;) I'm kicking things off with this message on the subject of STEP and ontologies ... which is actually material for several threads, so please feel free to go off on any tangent that appeals to you. One of my main interests in this regard is how best to integrate STEP and STEP-based ontologies into the development and application of the many other ontologies that are used in an enterprise. David Price's "exPRESS for free" experiments and his paper (see http://exff.org, and more specifically http://homepages.nildram.co.uk/~esukpc20/exff_r01a2/docs/semweb_step.html) have shown, among other things, that STEP AP's and instance data can be brought into the Semantic Web representational idiom (OWL, for example). Also, one of David's presentations at the recent ISO SC4 meeting, "Building a Bridge from STEP-land", lays out many of the issues I'd like to discuss here. [David, are you okay with making those slides either public or at least accessible to members of this list? Perhaps on the exff site? Or send it to me with your Wiki material ... hint, hint!] So here are some questions that naturally arise (this is not an exhaustive list -- please add to and comment on it!): (1) Would it be useful to define some "recommended practices" for doing (e.g.) STEP -> UML, STEP -> OWL, etc. transformations? (I think so.) If so, I propose to use this forum to discuss the development, prototyping, and testing of them. (2) Would it be useful to define some "high-level interfaces" for STEP-aware applications to use in interacting with STEP data? EuroSTEP -- and specifically, Ian Bailey :) -- has already done work in this area, which I think is a critical piece of the architecture needed to put access to STEP technology into the hands of enterprise developers. Again, I think the prototyping, testing, and validation of these interfaces could be a very appropriate activity for this forum. (3) Finally, I'd like to look at providing the developers of enterprise ontologies (which are proliferating like rabbits these days ;) with methods and tools to discover, integrate with, leverage, and avoid duplicating (!) STEP-based ontologies -- by which I mean to include both "canonically derived" ones in (1) and the possibly more "adapted" ones of (2). Of course, this issue is not unique to STEP, so there may be both COTS and open-source solutions out there, but the ones I've seen seem to be relatively primitive. Cheers, Steve From lubell at nist.gov Tue Apr 6 13:13:54 2004 From: lubell at nist.gov (Joshua Lubell) Date: Tue Apr 6 12:45:56 2004 Subject: [step-os] pyExpress Message-ID: <4072E552.7060003@nist.gov> Can somebody tell me the status of pyExpress, Hans Peter de Koning's Python EXPRESS tool? This was presented a year ago at the NASA-ESA Workshop on Aerospace Product Data Exchange, but a Google search on "pyexpress" yields little more than the PowerPoint slides presented at said workshop. The workshop presentation mentioned that pyExpress would eventually be published as open source. Thank you, Josh -- Josh Lubell National Institute of Standards and Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Gaithersburg MD 20899-8263, USA Phone: 1-301-975-3563, Email: lubell@nist.gov From Hans-Peter.de.Koning at esa.int Wed Apr 7 09:00:16 2004 From: Hans-Peter.de.Koning at esa.int (Hans-Peter.de.Koning@esa.int) Date: Wed Apr 7 08:32:52 2004 Subject: [step-os] pyExpress Message-ID: Joshua, I am also on this list (since Monday) so I'm probably the person to answer your question. Indeed pyExpress has not been published yet, because we are still waiting for a suitable formal ESA open source license. We have made considerable progress though, and I will report on this at the upcoming NASA-ESA Workshop on Product Data Exchange (at EADS/Astrium, Friedrichshafen, Germany, 20-23 April 2004, see www.estec.esa.int/conferences for details). The abstract for the presentation summarizes the status: -------------------------------------------------------------------------------- Abstract 6th NASA-ESA Workshop on Product Data Exchange "Developments with pyExpress - an open source Python binding" Hans Peter de Koning, ESA/ESTEC, Noordwijk, The Netherlands -------------------------------------------------------------------------------- pyExpress is a development by the European Space Agency (ESA) to create an open source software (OSS) toolkit for the efficient development and verification of ISO 10303 (STEP) based product data exchange protocols. It has the following objectives: - lower the threshold for implementation of STEP standards as much as possible; - ensure the availability of data exchange solutions in the long term; - avoid any dependency on third party run-time licenses; - ensure full portability across all computer platforms / compiler combinations used for (space) analysis tools; - provide an OSS environment for STEP which is comparable in power and effectiveness with XML OSS environments. pyExpress was started as an internal project at ESTEC in October 2002. It can now generate an efficient, early-binding Python class library from most EXPRESS (ISO 10303-11 edition 1) schemas. In conjunction with the module pyExpressLib.py it provides an easy to use API for the implementation of STEP data import and export interfaces. It contains complete reader and writer modules for ISO 10103-21 text encoding (a.k.a. 'Part 21 File' or 'STEP Physical File'). The API is an early-binding mapping of Python classes on EXPRESS ENTITYs and TYPEs. Because Python maps very well onto EXPRESS and also has a very elegant syntax as well as extremely powerful object-oriented reflection and text processing functionality, this leads to a very natural, easy-to-use programming interface. In order to support developers of data exchange interfaces who prefer using C++, also a C++ API generator in the form of an early-binding ANSI C++ class library has been implemented. This functionality is in advanced prototype stage. Recently ESA awarded a contract to the Computer Science Department of the University of Manchester (UoM) in the UK to merge the best of pyExpress with the UoM's Jex Java/EXPRESS toolsuite to create a very robust OSS EXPRESS Software Development Kit. pyExpress has not yet been published on the web, pending the completion of an approved ESA open source license. Publication is expected later in 2004. Furthermore we have used the Python libraries generated with pyExpress successfully to create TASverter, a command line tool that converts thermal-analysis models from/to ESARAD, THERMICA and STEP-TAS Part 21 files. Performance is quite good and the Python language maps very well onto EXPRESS, giving us a high productivity in creating the mapping logic. Large International Space Station models get converted in about half a minute on an average 1Ghz PC. Details at http://www.estec.esa.int/thermal/tools/tasverter.html Best regards, Hans Peter -- Hans Peter de Koning -- mailto:Hans-Peter.de.Koning@esa.int European Space Research and Technology Centre (ESTEC) Directorate of Technical and Quality Management Mechanical Engineering Department Thermal and Structures Division Analysis and Verification Section (ESA D/TEC-MCV) http://www.estec.esa.int/thermal/tools Joshua Lubell Sent by: step-os-bounces@ned.gsfc.nasa.gov 2004-04-06 19:13 Please respond to STEP Open-Source To: step-os@ned.gsfc.nasa.gov cc: Subject: [step-os] pyExpress Can somebody tell me the status of pyExpress, Hans Peter de Koning's Python EXPRESS tool? This was presented a year ago at the NASA-ESA Workshop on Aerospace Product Data Exchange, but a Google search on "pyexpress" yields little more than the PowerPoint slides presented at said workshop. The workshop presentation mentioned that pyExpress would eventually be published as open source. Thank you, Josh -- Josh Lubell National Institute of Standards and Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Gaithersburg MD 20899-8263, USA Phone: 1-301-975-3563, Email: lubell@nist.gov _______________________________________________ step-os mailing list step-os@step.nasa.gov http://step.nasa.gov/mailman/listinfo/step-os From david.d.briggs at boeing.com Wed Apr 7 13:49:19 2004 From: david.d.briggs at boeing.com (Briggs, David D) Date: Wed Apr 7 13:23:38 2004 Subject: [step-os] Open source for data retention Message-ID: Hello fellow Open Source advocates, First, thanks to Steve for setting up this mailing list. The more we encourage a community of open source thinkers, the better. I would like to add another flavor to the list of reasons why people are interested in open source for product data. My focus area is the long term retention of product data. Open formats, such as STEP, are excellent strategies for allowing the data to remain useful independent of the software system that created it. This independence of data is a key approach for the long term retention of data. Open formats allow the data to be understood years after it is created. Open source tools that operate on the data allow the utility of the data to be ensured as long as the source can be executed in some future computing environment. So my main interest in open source is the availability of tools that add value to the data, such as viewers, parsers, etc. Having these tools in an open source environment promises that the tools as well as the data are preserved in a way that will interpretable years in the future. I look forward to many interesting discussions on this list. David Briggs Associate Technical Fellow Boeing Commercial Airplanes, Information Systems P.O. Box 3707, M/S 2R-97 Seattle, WA 98124-2207 (206)544-5310 FAX: (206)544-5889 From Harold.P.Frisch at nasa.gov Mon Apr 19 10:14:03 2004 From: Harold.P.Frisch at nasa.gov (Harold P. Frisch) Date: Mon Apr 19 09:58:52 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application Message-ID: Anyone Some help would be appreciated. I've just made an attempt to use Protege to create an ontology based representation of an IDEF0 model for a model-intensive, model based systems engineering design process. I wanted to use this model both as a USE case to bound the AP233 "Model Management" capability that I'm trying to define and too also define an OWL model for the complex set of intertwined models and model parameters used in the design process. I also wanted to learn Protege via a NASA relevant problem that I understood. It didn't take me long to run into a brick wall. I can't figure out if the wall is real STEP-SysML-OWL problem or a product of my lack of Protege experience. A push over this wall would be appreciated. The attachment defines my problem and asks 4 questions that I can't answer. Thanks Harry -- Harry ----------- Harold P. Frisch Code 592 Systems Engineering Division NASA/Goddard Space Flight Center Greenbelt, MD 20771 (301) 286-8730 Voice (301) 286-0612 Fax harry.frisch@gsfc.nasa.gov or Harold.P.Frisch@nasa.gov If you're rejected by NASA's SPAM filter use: hpfrisch@pop700.gsfc.nasa.gov -------------- next part -------------- A non-text attachment was scrubbed... Name: IDEF0_protege_model.doc Type: application/msword Size: 119296 bytes Desc: not available Url : http://step.nasa.gov/pipermail/step-os/attachments/20040419/2e9956ba/IDEF0_protege_model-0001.doc From Harold.P.Frisch at nasa.gov Mon Apr 19 14:29:45 2004 From: Harold.P.Frisch at nasa.gov (Harold P. Frisch) Date: Mon Apr 19 14:00:33 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: References: Message-ID: Mike Thanks, you say it more clearly than Steve did. I see what your saying. I could never get totally comfortable with the notion of "class" or the reasons for the class inheritance caution flags that Mike D. and David O. kept waving. Now I understand. This is telling me that any OWL/STEP/SysML mapping for state and function based behavior could be non-trivial. Maybe someone has a simple solution to this problem and that its resolution is already built into STEP. I can't buy the notion that domains characterized by IDEF0 type activities and sub-activities are out of scope for OWL, Protege and ontologies. Harry >Hi Harry, > >I've not done anything with Protege, (just looked at the web pages briefly >to get a quick confirmation of what I thought it was about), so I may be >offbase, but here goes. > >I think your problem is a wrong viewpoint. Classes and subclasses are not >the same as activities and sub-activities in IDEF0. Classes have subclasses >that are related with the concept "is a", i.e. class "Author" has >subclasses "Columnist", "Editor", etc. (an example from the Protege web >site). The subclasses share ALL the characteristics of the superclasses, >and add some more (normally) to specialize the superclass. If you try to >use subclasses for the IDEF0 sub-activities you run into a number of >problems, since the sub-activities are not each more specialized versions >of the entire parent activity, they are each a PART of the parent activity, >all the sub-activities are added together to make the whole parent >activity. Same thing happens with the ICOMs, they add together in the >sub-levels to make composite flows at the parent level. One characteristic >of IDEF0 diagrams is that if you want you can plot the leaf nodes all on >one big sheet and collapse all the intermediate boxes out of the diagram >and lose nothing. If you do that with class diagrams you lose lots because >the inheritance from superclass to subclass carries all the common >attributes. > >This is probably the same thing Steve Waterbury observed as a "class >inheritance problem". > >Just my very uninformed observation, take it for what it's worth.... > > >--- >Mike Loeffler - Global IT Leader | voice: +1 (248) 685-6102 >GM Electrical, Controls & Software | fax: +1 (248) 685-5154 >3300 General Motors Rd. | mailto:michael.loeffler@gm.com >Milford, MI 48380-3726 | Mailstop 483-340-111 > > > > > "Harold P. Frisch" > step-os@ned.gsfc.nasa.gov > gov> cc: >sfriedenthal , Chris >Paredis > Sent by: >, David Oliver , >"Ives, Barry" > step-os-bounces@ned.gs , >jskipper@vitechcorp.com, Michael Dickerson > fc.nasa.gov Subject: >[step-os] Problems in applying Protege to IDEF0, with an >AP233 > >application > > > 04/19/2004 10:14 >AM > Please respond to >STEP > >Open-Source > > > > > > > > >Anyone > >Some help would be appreciated. > >I've just made an attempt to use Protege to create an ontology based >representation of an IDEF0 model for a model-intensive, model based >systems engineering design process. I wanted to use this model both >as a USE case to bound the AP233 "Model Management" capability that >I'm trying to define and too also define an OWL model for the complex >set of intertwined models and model parameters used in the design >process. I also wanted to learn Protege via a NASA relevant problem >that I understood. > >It didn't take me long to run into a brick wall. I can't figure out >if the wall is real STEP-SysML-OWL problem or a product of my lack of >Protege experience. > >A push over this wall would be appreciated. > >The attachment defines my problem and asks 4 questions that I can't answer. > >Thanks > >Harry > > > >-- >Harry >----------- >Harold P. Frisch >Code 592 >Systems Engineering Division >NASA/Goddard Space Flight Center >Greenbelt, MD 20771 >(301) 286-8730 Voice >(301) 286-0612 Fax >harry.frisch@gsfc.nasa.gov >or >Harold.P.Frisch@nasa.gov > >If you're rejected by NASA's SPAM filter use: >hpfrisch@pop700.gsfc.nasa.gov(See attached file: IDEF0_protege_model.doc) >_______________________________________________ >step-os mailing list >step-os@step.nasa.gov >http://step.nasa.gov/mailman/listinfo/step-os > > > > > >Attachment converted: Macintosh HD:IDEF0_protege_model 1.doc >(WDBN/MSWD) (0006B8F8) -- Harry ----------- Harold P. Frisch Code 592 Systems Engineering Division NASA/Goddard Space Flight Center Greenbelt, MD 20771 (301) 286-8730 Voice (301) 286-0612 Fax harry.frisch@gsfc.nasa.gov or Harold.P.Frisch@nasa.gov If you're rejected by NASA's SPAM filter use: hpfrisch@pop700.gsfc.nasa.gov From golux at comcast.net Mon Apr 19 16:35:39 2004 From: golux at comcast.net (Stephen C. Waterbury) Date: Mon Apr 19 16:07:16 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: References: Message-ID: <4084381B.6050505@comcast.net> Harold P. Frisch wrote: > This is telling me that any OWL/STEP/SysML mapping for state and > function based behavior could be non-trivial. > > Maybe someone has a simple solution to this problem and that its > resolution is already built into STEP. As you know, state and function are extremely general concepts that are represented with various flavors in various domains ... any useful unified representation will need to prove that it can be mapped to at least the most important domain-specific representations. I would *not* expect that to be non-trivial, and I doubt that anyone has a "simple" solution, although there are probably pieces of a solution available. I'd be surprised if the representation for them is even uniform *within* STEP -- for example, AP210 has a model for function that I doubt any other AP is aware of, and I'd be surprised if no other AP has one. > I can't buy the notion that domains characterized by IDEF0 type > activities and sub-activities are out of scope for OWL, Protege and > ontologies. Ontologies and processes are two different things, and there are different kinds of tools that are best adapted to modeling them. No *ontology* domain is out of scope for Protege. You can use Protege to model an ontology for processes -- meaning you can use it to define the types of objects that are used in IDEF0 ("process", "input", "output", etc.) and their properties and relationships. However, Protege is not suited to modeling how processes actually work -- that's what IDEF0 is (one of the methodologies) for. IDEF0 tools already have the IDEF0 ontology built into them, so they give you a "palette" with the IDEF0 classes -- process, input, output, etc. The difference is similar to the relationship between an EXPRESS modeling tool and a CAD tool. You can use the EXPRESS modeling tool to develop a model of the information that the CAD tool produces, but you wouldn't use an EXPRESS tool to do CAD modeling -- you use a CAD tool for that. :) One more note: although an ontology tool like Protege is not useful for modeling processes, that doesn't mean you can't use OWL to exchange IDEF0 models. OWL is very flexible in what it can represent, so you could use it to represent both an ontology and a set of instance data -- so for example you could put the IDEF0 ontology *and* an IDEF0 process model both into an OWL file. Mike wrote: >> [Mike's nice explanation omitted for brevity ...] >> This is probably the same thing Steve Waterbury observed as a "class >> inheritance problem". Yes, exactly. More generally, the process/subprocess relationship is quite different from the class/subclass relationship. Which is why I kind of cringed when you told me what you were going to attempt. ;) Cheers, Steve From golux at comcast.net Mon Apr 19 17:06:00 2004 From: golux at comcast.net (Stephen C. Waterbury) Date: Mon Apr 19 17:10:17 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: <4084381B.6050505@comcast.net> References: <4084381B.6050505@comcast.net> Message-ID: <40843F38.5000707@comcast.net> Stephen C. Waterbury wrote: > Ontologies and processes are two different things, and there > are different kinds of tools that are best adapted to modeling > them. Just to make sure this horse is really dead ... :) ... With tools, the main issue is *visualization*. So that summarizes the difference between Protege and IDEF0 tools: Protege lets you visualize an ontology (mainly, its class hierarchy and definitions). IDEF0 tools let you visualize a process. From Harold.P.Frisch at nasa.gov Mon Apr 19 20:11:16 2004 From: Harold.P.Frisch at nasa.gov (Harold P. Frisch) Date: Mon Apr 19 20:32:03 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application Message-ID: <1082419875.40846aa42a3bd@webpop.gsfc.nasa.gov> Steve We fail to have a common view of visualization and ontology. I looked at Protege and expected that it would help me visualize the activity hierarchy of the process and precisely define the data flows within the process. I expected to be able to define the semantics that defined the sub-activities within an activity breakdown and the semantics that defined sub-activity inputs, controls, mechanisms and outputs. For the last 40 years I've observed ill-defined activities with ill-defined data flows in a multi-disciplinary design environment with harmonization and common agreement achieved by chance encounter in the cafeteria. I was hoping that Protege and this ontology stuff would allow activity breakdowns and essential data flows to be well defined. Yes ontologies and processes are two different things; however, processes have semantics. I was hoping that a process ontology would enable one to: o obtain an explicit description of the process (concepts, properties and attributes of concepts, constraints on properties and attributes) o obtain a common vocabulary and shared understanding o obtain a definition of process terms and relations among them o arrange the concept is a hierarchy (activity, sub-activity) o define properties that an activity or sub-activity can have o share a common understanding of process information o enable reuse of process domain knowledge o make process assumptions explicit List stolen from slides 4-9 in "Ontology Engineering for the Semantic Web and Beyond" by Natalya F. Noy Stanford University and adapted to the process ontology problem. Sorry The old horse ain't dead yet. Harry Quoting "Stephen C. Waterbury" : > Stephen C. Waterbury wrote: > > > Ontologies and processes are two different things, and there > > are different kinds of tools that are best adapted to modeling > > them. > > Just to make sure this horse is really dead ... :) ... > > With tools, the main issue is *visualization*. So that > summarizes the difference between Protege and IDEF0 tools: > > Protege lets you visualize an ontology (mainly, its class > hierarchy and definitions). > > IDEF0 tools let you visualize a process. Actually I like to read the definitions of the arrows and boxes so that I can visualize the process ontology. The display is of secondary interest to me. > > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://step.nasa.gov/mailman/listinfo/step-os > Harry From golux at comcast.net Mon Apr 19 22:40:25 2004 From: golux at comcast.net (Stephen Waterbury) Date: Mon Apr 19 22:18:18 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: <1082419875.40846aa42a3bd@webpop.gsfc.nasa.gov> References: <1082419875.40846aa42a3bd@webpop.gsfc.nasa.gov> Message-ID: <40848D99.4020805@comcast.net> Before I answer Harry, let me invite all those whom he has cc'ed to join the "STEP Open-Source" list (step-os) if you are interested in discussions of STEP-related open-source and open-architecture issues. The list is open to the public, so all you have to do to subscribe is go to http://step.nasa.gov/mailman/listinfo/step-os Then we won't have to keep cc'ing you. :) Harold P. Frisch wrote: > We fail to have a common view of visualization and ontology. I looked at > Protege and expected that it would help me visualize the activity hierarchy of > the process and precisely define the data flows within the process. I expected > to be able to define the semantics that defined the sub-activities within an > activity breakdown and the semantics that defined sub-activity inputs, > controls, mechanisms and outputs. Actually, you *can* do that, but not the way you thought. :) > For the last 40 years I've observed ill-defined activities with ill-defined > data flows in a multi-disciplinary design environment with harmonization and > common agreement achieved by chance encounter in the cafeteria. > > I was hoping that Protege and this ontology stuff would allow activity > breakdowns and essential data flows to be well defined. Ontologies are part of the solution to that problem, just as STEP is part of the solution for an integrated engineering environment, but not the whole solution. > Yes ontologies and processes are two different things; however, processes have > semantics. ... and they have ontologies too. No contradiction there. :) > I was hoping that a process ontology would enable one to: ... Ontologies will indeed help with all the items in this list, but that doesn't mean that ontology-definition tools (such as Protege) are the best tools to use for *everything* related to each item. > o obtain an explicit description of the process (concepts, properties and > attributes of concepts, constraints on properties and attributes) > o obtain a common vocabulary and shared understanding > o obtain a definition of process terms and relations among them Protege can directly model all of these. > o arrange the concept [in] a hierarchy (activity, sub-activity) ^^^^^^^^^^^^^^^^^^^^^^ Ah, you're cheating Harry -- the slide actually said "subclass-superclass" hierarchy! And that's what it *meant*. "activity/sub-activity" is a completely different semantic. An activity and a sub-activity are both *instances* of a single class: the "activity" class. Their relationship to each other is similar to the relationship between parts in a STEP assembly -- in an IDEF0 tool, you can "drill down" into an activity to see the sub-activities of which it is composed, just as you can drill down into a part in an assembly browser to see the parts of which it is composed. But you're not looking at the subclasses of the assembled part -- just its components. In Protege, if you have define a class "activity", you might define subclasses of activity that are specific to the NASA mission lifecycle like "spacecraft assembly" = SA "instrument assembly" = IA "vibration testing" = VT So our class hierarchy is just one level deep: NASA Mission Activity <- class --------------------- / | \ SA IA VT <- subclasses Now let's define a NASA mission life cycle process template (or instance) -- it might be something like: (SA) <- instance of SA / \ (IA) (VT) (+ other stuff) <- instances of IA and VT, etc. / \ (VT) (+ other other stuff) <- instance of VT, etc. Notice that when an actual process or process template is defined, instances of the same activity (class) can happen at various places -- e.g., vibration testing (VT) is done at various levels. (Yes, you could define more precise subclasses of VT for these, say "Spacecraft Vibration Testing" (SVT) and "Instrument Vibration Testing" (IVT), as each one no doubt has unique properties, but that would prove my point even further: the new SVT and IVT classes would be subclasses of VT, but their instances would be sub-*activities* of SA and IA, respectively, so the diagrams -- and more importantly, their semantics -- would still be different.) Spacecraft Assembly, Instrument Assembly, and Vibration Testing are all subclasses of the general "NASA Mission Activity" class, and they are also sub-activities of NASA Mission Activity, but the definition of the NASA Mission Activity process does not necessarily assemble them (activity/sub-activity) in the same order as their class hierarchy (class/subclass). The *class* hierarchy (the first diagram) and the properties and definitions associated with the classes, is what Protege is designed to model, visualize, and define. The *process* hierarchy (the second diagram) is what IDEF0 and other process modeling technologies are designed to model, visualize, and produce. The ontology provides definitions and class relationships for the concepts used by the process model -- as you correctly inferred from the slides about ontologies. However, your reading of "subclass-superclass" as "activity, sub-activity" was not correct. > o define properties that an activity or sub-activity can have Yes -- this is meta-data again, which is what Protege does best. > o share a common understanding of process information Protege will help you define all the terms you use in defining processes, but it is not designed to help you model process templates or instances of processes -- process-modeling tools (such as IDEF0) and workflow modeling tools do that. > o enable reuse of process domain knowledge > o make process assumptions explicit Yep. > List stolen from slides 4-9 in "Ontology Engineering for the Semantic Web and > Beyond" by Natalya F. Noy Stanford University and adapted to the process > ontology problem. Very nice slides. > Sorry > The old horse ain't dead yet. It was not my intention to identify the horse with *you* ... ;) Cheers, Steve From golux at comcast.net Mon Apr 19 23:49:34 2004 From: golux at comcast.net (Stephen Waterbury) Date: Mon Apr 19 23:25:19 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: <1082419875.40846aa42a3bd@webpop.gsfc.nasa.gov> References: <1082419875.40846aa42a3bd@webpop.gsfc.nasa.gov> Message-ID: <40849DCE.8030508@comcast.net> Harry, I forgot to address your final remark ... see below. :) Harold P. Frisch wrote: >>Stephen C. Waterbury wrote: >>>Ontologies and processes are two different things, and there >>>are different kinds of tools that are best adapted to modeling >>>them. >> >>Just to make sure this horse is really dead ... :) ... >> >>With tools, the main issue is *visualization*. So that >>summarizes the difference between Protege and IDEF0 tools: >> >>Protege lets you visualize an ontology (mainly, its class >>hierarchy and definitions). >> >>IDEF0 tools let you visualize a process. > > Actually I like to read the definitions of the arrows and boxes so that I can > visualize the process ontology. The display is of secondary interest to me. Without the tool's visualization capabilities, you wouldn't have any arrows and boxes to read the definitions of -- you'd just be looking at a text (or binary) file. My point was simply that the central visual metaphor of a tool tells you what it has been designed to do: with Protege, it's classes; with IDEF0, it's activities. With Emacs or vi, it's code. :) A class icon in a Protege class diagram does share the "expansion" or "drill-down" behavior of an activity box in an IDEF0 diagram, but the meaning of the expansion (class/subclass vs. activity/sub-activity) is different. - Steve From Harold.P.Frisch at nasa.gov Tue Apr 20 07:51:04 2004 From: Harold.P.Frisch at nasa.gov (Harold P. Frisch) Date: Tue Apr 20 07:21:45 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: <007d01c42698$319e9f20$de02470a@TQRH> References: <007d01c42698$319e9f20$de02470a@TQRH> Message-ID: Ralph Many thanks - I suspected that you would have my answer. I'm having a problem opening the IDEF0.ppri file that you sent. Do I also need the IDEF0.pont and IDEF0.parc files? I had this problem with a file that I created a few days ago. It wouldn't open and I had to start over. I didn't loose anything of value so it wasn't a problem. I took a WORD look at the IDEF0.ppri file. We are using the same Protege version. The file content compared to that of the file that I developed is obviously different but they look very similar. I don't see any open (...)'s or other obvious flags. Any suggestions? Yes I'd like a copy of your paper. Harry At 10:27 PM -0700 4/19/04, rhodgson wrote: >Harry, > >I have a few Prot?g? models of process ontologies. I did my first >process modeling in Prot?g? back in my IBM days (around 1998) when >Prot?g? was much younger. Because of other needs I converted the models >into Prolog. I could send you a paper I wrote on that. > >Anyhow, I have atached a quick effort in OWL I did this evening to >express the basic constructs of IDEF0. The model has your example in it. >You were making an inappropriate use of subclasses. You will see what I >mean from the model I am sending. > >To model splits, forks and merge-splits etc., additional classes will be >needed. But the basic idea is the same. If you can say it in objects you >can say it in OWL (and more). > >On the topic of visualization, a constraint-based drawing system could >render IDEF0 models. This is the approach that cogito uses. > >Incidentally in the NASA Space Shuttle ontology we are modeling >processes. > >Ralph Hodgson >Executive Partner >TopQuadrant, Inc. >Office: (724) 846-9300 ext. 211, Direct: (703) 960-1028, Fax: (425) >955-5469, Cell: (781) 789-1664 >http://www.topquadrant.com > > > >-----Original Message----- >From: Harold P. Frisch [mailto:Harold.P.Frisch@nasa.gov] >Sent: Monday, April 19, 2004 9:15 AM >To: 'Ralph Hodgson' >Subject: Fwd: [step-os] Problems in applying Protege to IDEF0, with an >AP233 application > > >Ralph > >Forgot to put you on the cc list. If you've got any such Protege >examples that relate to my problem I'd appreciate a copy so that I >could see how you get around my brick wall. Glad to hear you closed >the loop with David Price. > >Harry > >>X-Sender: hpfrisch@pop700.gsfc.nasa.gov >>Date: Mon, 19 Apr 2004 10:14:03 -0400 >>To: step-os@ned.gsfc.nasa.gov >>From: "Harold P. Frisch" >>X-Mailman-Approved-At: Mon, 19 Apr 2004 09:58:49 -0400 >>Cc: sfriedenthal , >> Chris Paredis , >> David Oliver , "Ives, Barry" , >> jskipper@vitechcorp.com, Michael Dickerson >>Subject: [step-os] Problems in applying Protege to IDEF0, >> with an AP233 application >>X-BeenThere: step-os@step.nasa.gov >>X-Mailman-Version: 2.1.4 >>Reply-To: STEP Open-Source >>List-Id: STEP Open-Source >>List-Unsubscribe: , >> >>List-Archive: >>List-Post: >>List-Help: >>List-Subscribe: , >> >>Sender: step-os-bounces@ned.gsfc.nasa.gov >> >>Anyone >> >>Some help would be appreciated. >> >>I've just made an attempt to use Protege to create an ontology based >>representation of an IDEF0 model for a model-intensive, model based > >systems engineering design process. I wanted to use this model both >>as a USE case to bound the AP233 "Model Management" capability that >>I'm trying to define and too also define an OWL model for the >>complex set of intertwined models and model parameters used in the >>design process. I also wanted to learn Protege via a NASA relevant >>problem that I understood. >> >>It didn't take me long to run into a brick wall. I can't figure out >>if the wall is real STEP-SysML-OWL problem or a product of my lack >>of Protege experience. >> >>A push over this wall would be appreciated. >> >>The attachment defines my problem and asks 4 questions that I can't >>answer. >> >>Thanks >> >>Harry >> >> >> >>-- >>Harry >>----------- >>Harold P. Frisch >>Code 592 >>Systems Engineering Division >>NASA/Goddard Space Flight Center >>Greenbelt, MD 20771 >>(301) 286-8730 Voice >>(301) 286-0612 Fax >>harry.frisch@gsfc.nasa.gov >>or >>Harold.P.Frisch@nasa.gov >> >>If you're rejected by NASA's SPAM filter use: >>hpfrisch@pop700.gsfc.nasa.gov >> >> >> >>_______________________________________________ >>step-os mailing list >>step-os@step.nasa.gov http://step.nasa.gov/mailman/listinfo/step-os > > >-- >Harry >----------- >Harold P. Frisch >Code 592 >Systems Engineering Division >NASA/Goddard Space Flight Center >Greenbelt, MD 20771 >(301) 286-8730 Voice >(301) 286-0612 Fax >harry.frisch@gsfc.nasa.gov >or >Harold.P.Frisch@nasa.gov > >If you're rejected by NASA's SPAM filter use: >hpfrisch@pop700.gsfc.nasa.gov > > >Content-Type: application/octet-stream; > name="IDEF0.owl" >Content-Disposition: attachment; > filename="IDEF0.owl" > >Attachment converted: Macintosh HD:IDEF0.owl (????/----) (0006BA17) >Content-Type: application/octet-stream; > name="IDEF0.pprj" >Content-Disposition: attachment; > filename="IDEF0.pprj" > >Attachment converted: Macintosh HD:IDEF0.pprj (????/----) (0006BA18) -- Harry ----------- Harold P. Frisch Code 592 Systems Engineering Division NASA/Goddard Space Flight Center Greenbelt, MD 20771 (301) 286-8730 Voice (301) 286-0612 Fax harry.frisch@gsfc.nasa.gov or Harold.P.Frisch@nasa.gov If you're rejected by NASA's SPAM filter use: hpfrisch@pop700.gsfc.nasa.gov From rhodgson at topquadrant.com Tue Apr 20 11:11:19 2004 From: rhodgson at topquadrant.com (rhodgson) Date: Tue Apr 20 10:50:33 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application In-Reply-To: Message-ID: <004701c426e9$c01789a0$de02470a@TQRH> Harry, Strange that you cannot open the file. I just detatched the attachments to another place on my machine and loaded them fine. You only need the two files. Let me know if you have any more success. The Prolog paper is attached. You will see that I made some use of the Object Constraint Language to think through some of the process model constraints. Ralph Hodgson Executive Partner TopQuadrant, Inc. Office: (724) 846-9300 ext. 211, Direct: (703) 960-1028, Fax: (425) 955-5469, Cell: (781) 789-1664 http://www.topquadrant.com -----Original Message----- From: Harold P. Frisch [mailto:Harold.P.Frisch@nasa.gov] Sent: Tuesday, April 20, 2004 4:51 AM To: rhodgson@topquadrant.com Cc: STEP Open-Source; 'sfriedenthal'; 'Chris Paredis'; 'David Oliver'; 'Ives Barry'; jskipper@vitechcorp.com; 'Michael Dickerson'; 'Stephen Waterbury' Subject: RE: [step-os] Problems in applying Protege to IDEF0, with an AP233 application Ralph Many thanks - I suspected that you would have my answer. I'm having a problem opening the IDEF0.ppri file that you sent. Do I also need the IDEF0.pont and IDEF0.parc files? I had this problem with a file that I created a few days ago. It wouldn't open and I had to start over. I didn't loose anything of value so it wasn't a problem. I took a WORD look at the IDEF0.ppri file. We are using the same Protege version. The file content compared to that of the file that I developed is obviously different but they look very similar. I don't see any open (...)'s or other obvious flags. Any suggestions? Yes I'd like a copy of your paper. Harry At 10:27 PM -0700 4/19/04, rhodgson wrote: >Harry, > >I have a few Prot?g? models of process ontologies. I did my first >process modeling in Prot?g? back in my IBM days (around 1998) when >Prot?g? was much younger. Because of other needs I converted the models >into Prolog. I could send you a paper I wrote on that. > >Anyhow, I have atached a quick effort in OWL I did this evening to >express the basic constructs of IDEF0. The model has your example in >it. You were making an inappropriate use of subclasses. You will see >what I mean from the model I am sending. > >To model splits, forks and merge-splits etc., additional classes will >be needed. But the basic idea is the same. If you can say it in objects >you can say it in OWL (and more). > >On the topic of visualization, a constraint-based drawing system could >render IDEF0 models. This is the approach that cogito uses. > >Incidentally in the NASA Space Shuttle ontology we are modeling >processes. > >Ralph Hodgson >Executive Partner >TopQuadrant, Inc. >Office: (724) 846-9300 ext. 211, Direct: (703) 960-1028, Fax: (425) >955-5469, Cell: (781) 789-1664 http://www.topquadrant.com > > > >-----Original Message----- >From: Harold P. Frisch [mailto:Harold.P.Frisch@nasa.gov] >Sent: Monday, April 19, 2004 9:15 AM >To: 'Ralph Hodgson' >Subject: Fwd: [step-os] Problems in applying Protege to IDEF0, with an >AP233 application > > >Ralph > >Forgot to put you on the cc list. If you've got any such Protege >examples that relate to my problem I'd appreciate a copy so that I >could see how you get around my brick wall. Glad to hear you closed >the loop with David Price. > >Harry > >>X-Sender: hpfrisch@pop700.gsfc.nasa.gov >>Date: Mon, 19 Apr 2004 10:14:03 -0400 >>To: step-os@ned.gsfc.nasa.gov >>From: "Harold P. Frisch" >>X-Mailman-Approved-At: Mon, 19 Apr 2004 09:58:49 -0400 >>Cc: sfriedenthal , >> Chris Paredis , >> David Oliver , "Ives, Barry" , >> jskipper@vitechcorp.com, Michael Dickerson >> >>Subject: [step-os] Problems in applying Protege to IDEF0, >> with an AP233 application >>X-BeenThere: step-os@step.nasa.gov >>X-Mailman-Version: 2.1.4 >>Reply-To: STEP Open-Source >>List-Id: STEP Open-Source >>List-Unsubscribe: , >> >>List-Archive: >>List-Post: >>List-Help: >>List-Subscribe: , >> >>Sender: step-os-bounces@ned.gsfc.nasa.gov >> >>Anyone >> >>Some help would be appreciated. >> >>I've just made an attempt to use Protege to create an ontology based >>representation of an IDEF0 model for a model-intensive, model based > >systems engineering design process. I wanted to use this model both >>as a USE case to bound the AP233 "Model Management" capability that >>I'm trying to define and too also define an OWL model for the complex >>set of intertwined models and model parameters used in the design >>process. I also wanted to learn Protege via a NASA relevant problem >>that I understood. >> >>It didn't take me long to run into a brick wall. I can't figure out if >>the wall is real STEP-SysML-OWL problem or a product of my lack of >>Protege experience. >> >>A push over this wall would be appreciated. >> >>The attachment defines my problem and asks 4 questions that I can't >>answer. >> >>Thanks >> >>Harry >> >> >> >>-- >>Harry >>----------- >>Harold P. Frisch >>Code 592 >>Systems Engineering Division >>NASA/Goddard Space Flight Center >>Greenbelt, MD 20771 >>(301) 286-8730 Voice >>(301) 286-0612 Fax >>harry.frisch@gsfc.nasa.gov >>or >>Harold.P.Frisch@nasa.gov >> >>If you're rejected by NASA's SPAM filter use: >>hpfrisch@pop700.gsfc.nasa.gov >> >> >> >>_______________________________________________ >>step-os mailing list >>step-os@step.nasa.gov http://step.nasa.gov/mailman/listinfo/step-os > > >-- >Harry >----------- >Harold P. Frisch >Code 592 >Systems Engineering Division >NASA/Goddard Space Flight Center >Greenbelt, MD 20771 >(301) 286-8730 Voice >(301) 286-0612 Fax >harry.frisch@gsfc.nasa.gov >or >Harold.P.Frisch@nasa.gov > >If you're rejected by NASA's SPAM filter use: >hpfrisch@pop700.gsfc.nasa.gov > > >Content-Type: application/octet-stream; > name="IDEF0.owl" >Content-Disposition: attachment; > filename="IDEF0.owl" > >Attachment converted: Macintosh HD:IDEF0.owl (????/----) (0006BA17) >Content-Type: application/octet-stream; > name="IDEF0.pprj" >Content-Disposition: attachment; > filename="IDEF0.pprj" > >Attachment converted: Macintosh HD:IDEF0.pprj (????/----) (0006BA18) -- Harry ----------- Harold P. Frisch Code 592 Systems Engineering Division NASA/Goddard Space Flight Center Greenbelt, MD 20771 (301) 286-8730 Voice (301) 286-0612 Fax harry.frisch@gsfc.nasa.gov or Harold.P.Frisch@nasa.gov If you're rejected by NASA's SPAM filter use: hpfrisch@pop700.gsfc.nasa.gov -------------- next part -------------- A non-text attachment was scrubbed... Name: inap2000rh6.pdf Type: application/pdf Size: 105512 bytes Desc: not available Url : http://step.nasa.gov/pipermail/step-os/attachments/20040420/6b6377db/inap2000rh6-0001.pdf From Harold.P.Frisch at nasa.gov Thu Apr 22 13:03:13 2004 From: Harold.P.Frisch at nasa.gov (Harold P. Frisch) Date: Thu Apr 22 12:33:26 2004 Subject: [step-os] Problems in applying Protege to IDEF0, with an AP233 application Message-ID: >Date: Thu, 22 Apr 2004 10:54:58 -0400 >To: >From: "Harold P. Frisch" >Subject: RE: [step-os] Problems in applying >Protege to IDEF0, with an AP233 application >Cc: "'sfriedenthal'" >, "'Chris >Paredis'" , "'David >Oliver'" , "'Ives Barry'" >, , >"'Michael Dickerson'" , >"'Stephen Waterbury'" >Bcc: >X-Attachments: > >Ralph > >Very nice - I see what you've done. Now I'll >carry it a few levels of abstraction deeper to >see if I find anymore brick walls and too more >fully appreciate the IDEF0 capturing framework >that you have defined. > >Seems to me that a GIT or Univ of MD student >could use your framework to write a mapping tool >that would map an IDEF0 data file directly into >an OWL model. Is there some road block to this >that I don't see? > >All > >Ralph's earlier attachment only seems to work >with the most recent copy of Protege and the OWL >plugin. If you try it, keep a copy of the >original on the side so you can start over. >Whenever I do a SAVE I can't loose half the file >and can't recover. > >Harry > > > >At 10:27 PM -0700 4/19/04, rhodgson wrote: >>Harry, >> >>I have a few Prot?g? models of process ontologies. I did my first >>process modeling in Prot?g? back in my IBM days (around 1998) when >>Prot?g? was much younger. Because of other needs I converted the models >>into Prolog. I could send you a paper I wrote on that. >> >>Anyhow, I have atached a quick effort in OWL I did this evening to >>express the basic constructs of IDEF0. The model has your example in it. >>You were making an inappropriate use of subclasses. You will see what I >>mean from the model I am sending. >> >>To model splits, forks and merge-splits etc., additional classes will be >>needed. But the basic idea is the same. If you can say it in objects you >>can say it in OWL (and more). >> >>On the topic of visualization, a constraint-based drawing system could >>render IDEF0 models. This is the approach that cogito uses. >> >>Incidentally in the NASA Space Shuttle ontology we are modeling >>processes. >> >>Ralph Hodgson >>Executive Partner >>TopQuadrant, Inc. >>Office: (724) 846-9300 ext. 211, Direct: (703) 960-1028, Fax: (425) >>955-5469, Cell: (781) 789-1664 >>http://www.topquadrant.com >> >> >> >>-----Original Message----- >>From: Harold P. Frisch [mailto:Harold.P.Frisch@nasa.gov] >>Sent: Monday, April 19, 2004 9:15 AM >>To: 'Ralph Hodgson' >>Subject: Fwd: [step-os] Problems in applying Protege to IDEF0, with an >>AP233 application >> >> >>Ralph >> >>Forgot to put you on the cc list. If you've got any such Protege >>examples that relate to my problem I'd appreciate a copy so that I >>could see how you get around my brick wall. Glad to hear you closed >>the loop with David Price. >> >>Harry >> >>>X-Sender: hpfrisch@pop700.gsfc.nasa.gov >>>Date: Mon, 19 Apr 2004 10:14:03 -0400 >>>To: step-os@ned.gsfc.nasa.gov >>>From: "Harold P. Frisch" >>>X-Mailman-Approved-At: Mon, 19 Apr 2004 09:58:49 -0400 >>>Cc: sfriedenthal , >>> Chris Paredis , >>> David Oliver , "Ives, Barry" , >>> jskipper@vitechcorp.com, Michael Dickerson >>>Subject: [step-os] Problems in applying Protege to IDEF0, >>> with an AP233 application >>>X-BeenThere: step-os@step.nasa.gov >>>X-Mailman-Version: 2.1.4 >>>Reply-To: STEP Open-Source >>>List-Id: STEP Open-Source >>>List-Unsubscribe: , >>> >>>List-Archive: >>>List-Post: >>>List-Help: >>>List-Subscribe: , >>> >> >Sender: step-os-bounces@ned.gsfc.nasa.gov >>> >>>Anyone >>> >>>Some help would be appreciated. >>> >>>I've just made an attempt to use Protege to create an ontology based >>>representation of an IDEF0 model for a model-intensive, model based >> >systems engineering design process. I wanted to use this model both >>>as a USE case to bound the AP233 "Model Management" capability that >>>I'm trying to define and too also define an OWL model for the >>>complex set of intertwined models and model parameters used in the >>>design process. I also wanted to learn Protege via a NASA relevant >>>problem that I understood. >>> >>>It didn't take me long to run into a brick wall. I can't figure out >>>if the wall is real STEP-SysML-OWL problem or a product of my lack >>>of Protege experience. >>> >>>A push over this wall would be appreciated. >>> >>>The attachment defines my problem and asks 4 questions that I can't >>>answer. >>> >>>Thanks >>> >>>Harry >>> >>> >>> >>>-- >>>Harry >>>----------- >>>Harold P. Frisch >>>Code 592 >>>Systems Engineering Division >>>NASA/Goddard Space Flight Center >>>Greenbelt, MD 20771 >>>(301) 286-8730 Voice >>>(301) 286-0612 Fax >>>harry.frisch@gsfc.nasa.gov >>>or >>>Harold.P.Frisch@nasa.gov >>> >>>If you're rejected by NASA's SPAM filter use: >>>hpfrisch@pop700.gsfc.nasa.gov >>> >>> >>> >>>_______________________________________________ >>>step-os mailing list >>>step-os@step.nasa.gov http://step.nasa.gov/mailman/listinfo/step-os >> >> >>-- >>Harry >>----------- >>Harold P. Frisch >>Code 592 >>Systems Engineering Division >>NASA/Goddard Space Flight Center >>Greenbelt, MD 20771 >>(301) 286-8730 Voice >>(301) 286-0612 Fax >>harry.frisch@gsfc.nasa.gov >>or >>Harold.P.Frisch@nasa.gov >> >>If you're rejected by NASA's SPAM filter use: >>hpfrisch@pop700.gsfc.nasa.gov >> >> >>Content-Type: application/octet-stream; >> name="IDEF0.owl" >>Content-Disposition: attachment; >> filename="IDEF0.owl" >> >>Attachment converted: Macintosh HD:IDEF0.owl (????/----) (0006BA17) >>Content-Type: application/octet-stream; >> name="IDEF0.pprj" >>Content-Disposition: attachment; >> filename="IDEF0.pprj" >> >>Attachment converted: Macintosh HD:IDEF0.pprj (????/----) (0006BA18) > > >-- > >Harry >----------- >Harold P. Frisch >Code 592 >Systems Engineering Division >NASA/Goddard Space Flight Center >Greenbelt, MD 20771 >(301) 286-8730 Voice >(301) 286-0612 Fax >harry.frisch@gsfc.nasa.gov >or >Harold.P.Frisch@nasa.gov > >If you're rejected by NASA's SPAM filter use: >hpfrisch@pop700.gsfc.nasa.gov -- Harry ----------- Harold P. Frisch Code 592 Systems Engineering Division NASA/Goddard Space Flight Center Greenbelt, MD 20771 (301) 286-8730 Voice (301) 286-0612 Fax harry.frisch@gsfc.nasa.gov or Harold.P.Frisch@nasa.gov If you're rejected by NASA's SPAM filter use: hpfrisch@pop700.gsfc.nasa.gov -------------- next part -------------- An HTML attachment was scrubbed... URL: http://step.nasa.gov/pipermail/step-os/attachments/20040422/916bad11/attachment.html From david.price at eurostep.com Mon Apr 26 12:20:53 2004 From: david.price at eurostep.com (David Price) Date: Mon Apr 26 11:50:57 2004 Subject: [step-os] And the kick-off: STEP and ontologies In-Reply-To: <4071E3E8.3@comcast.net> Message-ID: <001f01c42baa$723977a0$2101a8c0@esukpc20> FYI, Rob B and I will be writing a more formal paper to explain the continued work we've been doing for STEP-based data exchange and OWL. This will address your question 3 a bit too. This is, however, not the same thing as a STEP-based ontology. As for STEP-OWL, this isn't just EXPRESS->OWL. I think there will have to be AP-specific translators to turn STEP exchange files into ontology-driven data. This is because some of the instances in the STEP file represent classes in the ontology. As for STEP->UML, Part 25 is defined for EXPRESS->UML if that's what you mean. It really needs an edition 2 to cover EXPRESS 2, UML Profiles, etc. More to come... David > -----Original Message----- > From: step-os-bounces@ned.gsfc.nasa.gov [mailto:step-os- > bounces@ned.gsfc.nasa.gov] On Behalf Of Stephen C. Waterbury > Sent: 05 April 2004 23:56 > To: STEP Open-Source > Subject: [step-os] And the kick-off: STEP and ontologies > > Greetings, fellow STEP-OSers! > > Judging by our membership (which you can see if you go > to http://step.nasa.gov/mailman/listinfo/step-os and enter > your address and password), I think we have a quorum. > Also, all messages to this list go into our archive: > http://www.step.nasa.gov/pipermail/step-os ... so late-comers > will have access to our discussion history, anyway. ;) > > I'm kicking things off with this message on the subject of > STEP and ontologies ... which is actually material for > several threads, so please feel free to go off on any > tangent that appeals to you. > > One of my main interests in this regard is how best to integrate > STEP and STEP-based ontologies into the development and application > of the many other ontologies that are used in an enterprise. > > David Price's "exPRESS for free" experiments and his paper > (see http://exff.org, and more specifically > http://homepages.nildram.co.uk/~esukpc20/exff_r01a2/docs/semweb_step.html) > have shown, among other things, that STEP AP's and instance data > can be brought into the Semantic Web representational idiom > (OWL, for example). > > Also, one of David's presentations at the recent ISO SC4 > meeting, "Building a Bridge from STEP-land", lays out many > of the issues I'd like to discuss here. [David, are you okay > with making those slides either public or at least accessible > to members of this list? Perhaps on the exff site? Or send > it to me with your Wiki material ... hint, hint!] > > So here are some questions that naturally arise (this is not > an exhaustive list -- please add to and comment on it!): > > (1) Would it be useful to define some "recommended practices" for > doing (e.g.) STEP -> UML, STEP -> OWL, etc. transformations? > (I think so.) If so, I propose to use this forum to discuss > the development, prototyping, and testing of them. > > (2) Would it be useful to define some "high-level interfaces" for > STEP-aware applications to use in interacting with STEP data? > EuroSTEP -- and specifically, Ian Bailey :) -- has already done > work in this area, which I think is a critical piece of the > architecture needed to put access to STEP technology into the > hands of enterprise developers. Again, I think the prototyping, > testing, and validation of these interfaces could be a very > appropriate activity for this forum. > > (3) Finally, I'd like to look at providing the developers > of enterprise ontologies (which are proliferating like rabbits > these days ;) with methods and tools to discover, integrate with, > leverage, and avoid duplicating (!) STEP-based ontologies -- by > which I mean to include both "canonically derived" ones in (1) > and the possibly more "adapted" ones of (2). Of course, this > issue is not unique to STEP, so there may be both COTS and > open-source solutions out there, but the ones I've seen seem > to be relatively primitive. > > Cheers, > Steve > > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://step.nasa.gov/mailman/listinfo/step-os From golux at comcast.net Mon Apr 26 13:03:01 2004 From: golux at comcast.net (Stephen C. Waterbury) Date: Mon Apr 26 12:33:09 2004 Subject: [step-os] And the kick-off: STEP and ontologies In-Reply-To: <001f01c42baa$723977a0$2101a8c0@esukpc20> References: <001f01c42baa$723977a0$2101a8c0@esukpc20> Message-ID: <408D40C5.5080109@comcast.net> David Price wrote: > FYI, Rob B and I will be writing a more formal paper to explain the > continued work we've been doing for STEP-based data exchange and OWL. This > will address your question 3 a bit too. This is, however, not the same thing > as a STEP-based ontology. Actually I've used the term "STEP-based ontology" pretty loosely. We might want to give it a more rigorous definition if we want to use it to classify ontologies. :) > As for STEP-OWL, this isn't just EXPRESS->OWL. I think there will have to be > AP-specific translators to turn STEP exchange files into ontology-driven > data. ... We are on exactly the same page here. > ... This is because some of the instances in the STEP file represent > classes in the ontology. Which probably also implies that we will need OWL Full. I expected that. > As for STEP->UML, Part 25 is defined for EXPRESS->UML if that's what you > mean. It really needs an edition 2 to cover EXPRESS 2, UML Profiles, etc. Yes ... there's more work ahead. :) Cheers, Steve From david.price at eurostep.com Mon Apr 26 13:23:28 2004 From: david.price at eurostep.com (David Price) Date: Mon Apr 26 12:54:47 2004 Subject: [step-os] And the kick-off: STEP and ontologies In-Reply-To: <408D40C5.5080109@comcast.net> Message-ID: <002a01c42bb3$304e8890$2101a8c0@esukpc20> Steve, ON: >> ... This is because some of the instances in the STEP file represent >> classes in the ontology. >Which probably also implies that we will need OWL Full. >I expected that. I don't know that this is the case yet. OWL Full would will only be required if the same class must also be treated as an individual *in the ontology*, it's not related to whether it was an entity instance in the EXPRESS. Whether we need class of class is the big question. For example, the entity product_category is likely a class of class, but we could simply take the instances of product_category and make them subclasses of a class called product. Using that approach, product_category doesn't really appear in the resulting ontology. On definitions, we've recognized two kinds of ISO 10303-related ontologies at the moment: 1) those used as reference data for a data exchange and 2) those that enable inferencing/reasoners. I guess we could also just use native OWL as an XML data exchange syntax by mapping EXPRESS entity to OWL class and entity instance to OWL individual as well. In that case, OWL replaces the Part 28 XML. So, there may be a third possibility worth discussing. Cheers, David > -----Original Message----- > From: step-os-bounces@ned.gsfc.nasa.gov [mailto:step-os- > bounces@ned.gsfc.nasa.gov] On Behalf Of Stephen C. Waterbury > Sent: 26 April 2004 18:03 > To: STEP Open-Source > Subject: Re: [step-os] And the kick-off: STEP and ontologies > > David Price wrote: > > FYI, Rob B and I will be writing a more formal paper to explain the > > continued work we've been doing for STEP-based data exchange and OWL. > This > > will address your question 3 a bit too. This is, however, not the same > thing > > as a STEP-based ontology. > > Actually I've used the term "STEP-based ontology" pretty > loosely. We might want to give it a more rigorous definition > if we want to use it to classify ontologies. :) > > > As for STEP-OWL, this isn't just EXPRESS->OWL. I think there will have > to be > > AP-specific translators to turn STEP exchange files into ontology-driven > > data. ... > > We are on exactly the same page here. > > > ... This is because some of the instances in the STEP file represent > > classes in the ontology. > > Which probably also implies that we will need OWL Full. > I expected that. > > > As for STEP->UML, Part 25 is defined for EXPRESS->UML if that's what you > > mean. It really needs an edition 2 to cover EXPRESS 2, UML Profiles, > etc. > > Yes ... there's more work ahead. :) > > Cheers, > Steve > > _______________________________________________ > step-os mailing list > step-os@step.nasa.gov > http://step.nasa.gov/mailman/listinfo/step-os From golux at comcast.net Mon Apr 26 14:26:28 2004 From: golux at comcast.net (Stephen C. Waterbury) Date: Mon Apr 26 13:56:26 2004 Subject: [step-os] And the kick-off: STEP and ontologies In-Reply-To: <002a01c42bb3$304e8890$2101a8c0@esukpc20> References: <002a01c42bb3$304e8890$2101a8c0@esukpc20> Message-ID: <408D5454.6040902@comcast.net> David Price wrote: ON: >>>... This is because some of the instances in the STEP file represent >>>classes in the ontology. > >>Which probably also implies that we will need OWL Full. >>I expected that. > > I don't know that this is the case yet. OWL Full would will only be required > if the same class must also be treated as an individual *in the ontology*, > it's not related to whether it was an entity instance in the EXPRESS. > Whether we need class of class is the big question. Good point (note that I said "probably" :). It could be implemented either way, or both ways. In any case, it might be possible to define an OWL-Full scope with well-defined sub-scopes that are OWL-DL, somewhat analogous to the late-bound and early-bound exchange formats, but potentially more coherent, I think. > For example, the entity > product_category is likely a class of class, but we could simply take the > instances of product_category and make them subclasses of a class called > product. Using that approach, product_category doesn't really appear in the > resulting ontology. Yep. That is probably the simplest approach, so it's probably the best for prototyping and getting something that works. More than that might be YAGNI, anyway. :) Of course, the ontologies that applications use internally won't necessarily be the same as the external exchange or API ontologies that they understand. (I'm still working on giving PanGalactic an internal meta-architecture along the lines of PEAK's MOF-like approach, so that it will be able to adapt to any species of OWL. :) > On definitions, we've recognized two kinds of ISO 10303-related ontologies > at the moment: 1) those used as reference data for a data exchange and 2) > those that enable inferencing/reasoners. That seems like a useful distinction. A third might be ARM-level, API-like ontologies. Analogous to the PDM Enablers, and related to (1) in a similar way, but more likely to be implemented in a Web Services context. > I guess we could also just use > native OWL as an XML data exchange syntax by mapping EXPRESS entity to OWL > class and entity instance to OWL individual as well. In that case, OWL > replaces the Part 28 XML. So, there may be a third possibility worth > discussing. Okay, a fourth. :) This worries me a little about the possible proliferation of "standard" exchange syntaxes, given the amount of work that has already gone into Part 28. On the one hand, well-defined conversions between them could be specified, and we could provide free tools to do that. But that is another dependency and would make it difficult to create a "simple" application interface that can deal with STEP data if there are so many formats in which it might receive it. But perhaps that's too much to hope for. Ah, for the good old days when there was just Part 21. Life was simple then! (Ha. :) Cheers, Steve