sigmakee-sumo Mailing List for Sigma
Status: Beta
Brought to you by:
apease
Menu
▾
▴
sigmakee-sumo — Discussion about inferencing on SUMO in Sigma, as well as SUMO in general.
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
(45) |
Apr
(22) |
May
(9) |
Jun
|
Jul
(2) |
Aug
(2) |
Sep
(2) |
Oct
|
Nov
(2) |
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
(3) |
Feb
(1) |
Mar
(1) |
Apr
(6) |
May
(10) |
Jun
(1) |
Jul
|
Aug
(1) |
Sep
|
Oct
(4) |
Nov
(2) |
Dec
|
| 2007 |
Jan
(1) |
Feb
(4) |
Mar
|
Apr
|
May
|
Jun
|
Jul
(5) |
Aug
|
Sep
|
Oct
(1) |
Nov
(2) |
Dec
|
| 2012 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(9) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2013 |
Jan
(1) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
| 2015 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
|
From: Adam P. <ada...@at...> - 2015-10-04 16:06:49
|
Hi Folks, SourceForge has had some stability problems, and the world seems to be moving to Git, so we've migrated the site to GitHub. We'll keep the old SourceForge site up, read-only, at least for a while. There might be some issues in the transition, so please let the group know if you encounter problems. The project remains open source, of course. Adam http://ontologyportal.github.io/sigmakee/ -- ------------------- Adam Pease http://www.ontologyportal.org http://www.articulatesoftware.com http://www.adampease.org |
|
From: Adam P. <ada...@at...> - 2013-10-05 04:03:43
|
Hi Folks, I've posted a new podcast, talking through how to define a new term in SUMO. Maybe it will be of interest. http://www.ontologyportal.org/SumoPodcast1.mp3 all the best, Adam -- ------------------- Adam Pease http://www.ontologyportal.org http://www.articulatesoftware.com http://www.adampease.org |
|
From: Ron B. <ron...@ya...> - 2013-01-31 08:07:47
|
hello! http://www.aesengineering.com.au/components/com_content/mainx.php?e8au9=0d3u Ron Bogdanoff 1/31/2013 9:07:39 AM |
|
From: Rich M. <rd...@cf...> - 2012-08-28 18:10:42
|
I recently attended a seminar on ontology creation, led by Adam Pease. For my class exercise, I looked at Trimble's 3D Warehouse: http://sketchup.google.com/3dwarehouse/ This site contains millions of SketchUp Models, uploaded by volunteers and (in a few cases) organizations. It is organized into large numbers (thousands?) of Collections, eg: http://sketchup.google.com/3dwarehouse/cldetails?mid=61a20f92b4d71a00567f368a407ede14&prevstart=0 http://sketchup.google.com/3dwarehouse/search?q=Brizo&styp=c&scoring=t&btnG=Search&reps=1 Using the search tool, it is possible to find Collection and Model pages which have particular tags, etc. However, there are no categories, let alone any ontology-based navigation. The collections may have to do with the nature of the models or merely reflect the interests of a user. The tags are simply words, so they may have multiple, conflicting meanings. Adam and I will be getting together to brainstorm on how ontologies could be created and used for this and similar (eg, eBay) sites. If this sounds interesting, please join us. The venue is a small Chinese restaurant that is just off the San Bruno Avenue exit from I280: SUMO workshop Wednesday, 8/29, 6-8 pm AU's Kitchen Bayhill Shopping Center San Bruno Avenue at Cherry San Bruno, CA http://www.auskitchen.us/index.html RSVP: Please send me a note (off-list) if you expect to attend. -r -- http://www.cfcl.com/rdm Rich Morin http://www.cfcl.com/rdm/resume rd...@cf... http://www.cfcl.com/rdm/weblog +1 650-873-7841 Software system design, development, and documentation |
|
From: Rich M. <rd...@cf...> - 2012-08-22 04:07:25
|
This seems somewhat relevant to SUMO: From Words to Concepts and Back: Dictionaries for Linking Text, Entities and Ideas http://googleresearch.blogspot.com/2012/05/from-words-to-concepts-and-back.html -r -- http://www.cfcl.com/rdm Rich Morin http://www.cfcl.com/rdm/resume rd...@cf... http://www.cfcl.com/rdm/weblog +1 650-873-7841 Software system design, development, and documentation |
|
From: Adam P. <ada...@at...> - 2012-08-18 23:50:47
|
Hi Folks, I'm going to be in San Jose on the evening of Aug 29. Let me know if you're interested in getting together for a SUMO/Sigma user group meeting. Adam -- ------------------------------- Adam Pease http://www.adampease.org NOTE: ada...@ea... will soon no longer be active Please use ap...@ar... |
|
From: Adam P. <ada...@at...> - 2012-08-16 00:20:56
|
Hi Rich, I think this is a good plan. One way forward is to try to get people using the tags even though they're not ontologists creating new tags. I'd hope that many of the 3D models are of the same ontological category - brown table, shaker table, dining room table etc. so even just tagging them all as SUMO's Table would provide a lot of leverage. Adam On Wed, 2012-08-15 at 17:09 -0700, Rich Morin wrote: > On Aug 15, 2012, at 16:35, Adam Pease wrote: > > I think you could create (or locate) SUMO concepts for the models > > posted at http://sketchup.google.com/3dwarehouse/ . For example, for > > the items I see on that page we have ReligiousBuilding, Spacecraft, > > FloweringPlant, and Piano. We need a new term Lavender to offer a > > better mapping than FloweringPlant. One could go through the repository > > concept by concept, creating new terms in SUMO, and mapping them to the > > warehouse with a &%synonymousExternalConcept statement. > > > Thanks for the hints! Problem is, there are thousands of collections (at > least) and millions of models. So, a brute-force approach is not going > to be feasible (assuming only a minute per model, categorizing three > million models would take about 25 person-years to process). > > However, a slightly more sophisticated approach might be feasible: > > * Start with Collections (fewer to work with, topically related) > * Do a first-order match, add some terms, rinse, repeat. > * Use Collection membership to inform Model categorization. > > -r > > -- > http://www.cfcl.com/rdm Rich Morin > http://www.cfcl.com/rdm/resume rd...@cf... > http://www.cfcl.com/rdm/weblog +1 650-873-7841 > > Software system design, development, and documentation > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Sigmakee-sumo mailing list > Sig...@li... > https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo -- ------------------------------- Adam Pease http://www.adampease.org NOTE: ada...@ea... will soon no longer be active Please use ap...@ar... |
|
From: Rich M. <rd...@cf...> - 2012-08-16 00:09:31
|
On Aug 15, 2012, at 16:35, Adam Pease wrote: > I think you could create (or locate) SUMO concepts for the models > posted at http://sketchup.google.com/3dwarehouse/ . For example, for > the items I see on that page we have ReligiousBuilding, Spacecraft, > FloweringPlant, and Piano. We need a new term Lavender to offer a > better mapping than FloweringPlant. One could go through the repository > concept by concept, creating new terms in SUMO, and mapping them to the > warehouse with a &%synonymousExternalConcept statement. Thanks for the hints! Problem is, there are thousands of collections (at least) and millions of models. So, a brute-force approach is not going to be feasible (assuming only a minute per model, categorizing three million models would take about 25 person-years to process). However, a slightly more sophisticated approach might be feasible: * Start with Collections (fewer to work with, topically related) * Do a first-order match, add some terms, rinse, repeat. * Use Collection membership to inform Model categorization. -r -- http://www.cfcl.com/rdm Rich Morin http://www.cfcl.com/rdm/resume rd...@cf... http://www.cfcl.com/rdm/weblog +1 650-873-7841 Software system design, development, and documentation |
|
From: Adam P. <ada...@at...> - 2012-08-15 23:35:47
|
Hi Rich, I think you could create (or locate) SUMO concepts for the models posted at http://sketchup.google.com/3dwarehouse/ . For example, for the items I see on that page we have ReligiousBuilding, Spacecraft, FloweringPlant, and Piano. We need a new term Lavender to offer a better mapping than FloweringPlant. One could go through the repository concept by concept, creating new terms in SUMO, and mapping them to the warehouse with a &%synonymousExternalConcept statement. Adam On Wed, 2012-08-15 at 16:26 -0700, Rich Morin wrote: > I didn't get very far on my class project, but I'm still quite > interested in getting feedback on how to approach it, etc. If > any of the following sounds interesting, please get in touch! > > -r > > > Basically, the question is > > How can ontological analysis and/or tools be used > to improve the Trimble (SketchUp) 3D Warehouse > (http://sketchup.google.com/3dwarehouse/)? > > I should note that I have no real say in how the 3D Warehouse > develops, but Trimble has said that they intend to rework things, > so they might be open to offbeat ideas. > > > At present, the 3D Warehouse has a fairly simple structure: > > Structure > ========= > > Warehouse > > There is only one 3D Warehouse. It contains millions of Models, > organized as Collections of Collections and Collections of Models. > > Collection > > A Collection may contain either Collections or Models, but not > both. A Collection may be contained in multiple Collections. > So, Collections form a directed graph, not a hierarchy. > > The 3DW maintains (for each Collection): > - a unique, opaque ID > - title, tags, and description > - the Owner (who created it) > - ratings and reviews > - which Collection(s) contain it > - which Collections (or Models) it contains > - when it was created and last modified > > Model > > A Model is a file containing the description of some number of 3D > entities (eg, Components, Edges, Faces, Groups). Each Model also > contains some amount of metadata, stored in attribute dictionaries. > > The 3DW maintains (for each Model): > - a unique, opaque ID > - file type (Collada, Google Earth, SketchUp) > - title, tags, and description > - ratings and reviews > - related Models (Other models you might like) > - the Author (who uploaded it) > - the Complexity (number of polygons) > - which Collection(s) contain it > - whether it has Dynamic Attributes > - whether (and where) it is Geolocated > - when it was created and last modified > - how many Views, Downloads, and Google Earth model clicks > it has had (over time) > > Author > > An Author is someone who has uploaded one or more Models. > > > Notions > ======= > > Use existing tags and harvest other ones from the title, description, etc. > Attempt to ground these tags in SUMO, possibly using some sort of Bayesian > techniques to disambiguate word senses. > > Use Geolocation information, if available, to characterize locations. > > Create an ontology-based browser for authors and users, allowing them to > specify unambiguous categories, etc. > > -- > http://www.cfcl.com/rdm Rich Morin > http://www.cfcl.com/rdm/resume rd...@cf... > http://www.cfcl.com/rdm/weblog +1 650-873-7841 > > Software system design, development, and documentation > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Sigmakee-sumo mailing list > Sig...@li... > https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo -- ------------------------------- Adam Pease http://www.adampease.org NOTE: ada...@ea... will soon no longer be active Please use ap...@ar... |
|
From: Adam P. <ada...@at...> - 2012-08-15 23:29:33
|
Hi Rich, Many thanks for the kind review! I agree with your summary about the mailing lists, and I like your top-level list of projects. Don't let me discourage you from pursuing namespace stuff! As you say, that's the beauty of open source. It may have more value than I think. The only thing I'd recommend is to spend some time creating a significant amount of ontology content before you decide where to put effort in expanding the tool set. Here's a list of project ideas I had on one slide - Tree-based editor for creating subclass/instance hierarchies in KIF - text editor integration for SUO-KIF/Sigma - Integrate E - Fix the set and list theories to conform to Zeno-Frankel or other major approach - Elaborate the case role structure, revising use of patient to be more specific - Package Sigma for all three four (or more) OS's so it's easy to install! - Reload WordNet files button - Allow loading of multiple wordnets - fix the error message reporting - help me work on a better manual! Adam On Wed, 2012-08-15 at 16:14 -0700, Rich Morin wrote: > I enjoyed Adam's course a great deal and hope to use some of > what I learned in both personal and paid projects. I'm also > hoping that Adam will set up a regular (eg, monthly) meeting > where folks can touch base, work on problems together, etc. > > > I'm not exactly sure about the boundaries between the -sumo > and -develop lists, but my take is that -sumo covers pretty > much everything except development topics on specific code > (eg, Sigma, CELT). Please forgive (and correct) any errors. > > > Near the end of the course, Adam put up a wish list of areas > where he could use some help. I don't have a copy, but here > are some general areas that subsume most of his items (and a > few that he didn't mention): > > community building, documentation, etc. > software packaging and installation > namespace management, syntactic macros, etc. > visualization and user experience > > Some of these areas are already being addressed; others are > longer-term and/or highly speculative in nature. In fact, > Adam seems quite skeptical about the utility of namespace > management, syntactic macros, etc. However, one of the joys > of Open Source is the chance to try out partly-baked ideas. > > Speaking of which, a couple of us are interested in getting > Sigma (etc) running under Clojure, then using that environment > to start playing with some ideas. Please get in touch if this > sounds intriguing... > > -r > > -- > http://www.cfcl.com/rdm Rich Morin > http://www.cfcl.com/rdm/resume rd...@cf... > http://www.cfcl.com/rdm/weblog +1 650-873-7841 > > Software system design, development, and documentation > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Sigmakee-sumo mailing list > Sig...@li... > https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo -- ------------------------------- Adam Pease http://www.adampease.org NOTE: ada...@ea... will soon no longer be active Please use ap...@ar... |
|
From: Rich M. <rd...@cf...> - 2012-08-15 23:26:20
|
I didn't get very far on my class project, but I'm still quite interested in getting feedback on how to approach it, etc. If any of the following sounds interesting, please get in touch! -r Basically, the question is How can ontological analysis and/or tools be used to improve the Trimble (SketchUp) 3D Warehouse (http://sketchup.google.com/3dwarehouse/)? I should note that I have no real say in how the 3D Warehouse develops, but Trimble has said that they intend to rework things, so they might be open to offbeat ideas. At present, the 3D Warehouse has a fairly simple structure: Structure ========= Warehouse There is only one 3D Warehouse. It contains millions of Models, organized as Collections of Collections and Collections of Models. Collection A Collection may contain either Collections or Models, but not both. A Collection may be contained in multiple Collections. So, Collections form a directed graph, not a hierarchy. The 3DW maintains (for each Collection): - a unique, opaque ID - title, tags, and description - the Owner (who created it) - ratings and reviews - which Collection(s) contain it - which Collections (or Models) it contains - when it was created and last modified Model A Model is a file containing the description of some number of 3D entities (eg, Components, Edges, Faces, Groups). Each Model also contains some amount of metadata, stored in attribute dictionaries. The 3DW maintains (for each Model): - a unique, opaque ID - file type (Collada, Google Earth, SketchUp) - title, tags, and description - ratings and reviews - related Models (Other models you might like) - the Author (who uploaded it) - the Complexity (number of polygons) - which Collection(s) contain it - whether it has Dynamic Attributes - whether (and where) it is Geolocated - when it was created and last modified - how many Views, Downloads, and Google Earth model clicks it has had (over time) Author An Author is someone who has uploaded one or more Models. Notions ======= Use existing tags and harvest other ones from the title, description, etc. Attempt to ground these tags in SUMO, possibly using some sort of Bayesian techniques to disambiguate word senses. Use Geolocation information, if available, to characterize locations. Create an ontology-based browser for authors and users, allowing them to specify unambiguous categories, etc. -- http://www.cfcl.com/rdm Rich Morin http://www.cfcl.com/rdm/resume rd...@cf... http://www.cfcl.com/rdm/weblog +1 650-873-7841 Software system design, development, and documentation |
|
From: Rich M. <rd...@cf...> - 2012-08-15 23:14:40
|
I enjoyed Adam's course a great deal and hope to use some of what I learned in both personal and paid projects. I'm also hoping that Adam will set up a regular (eg, monthly) meeting where folks can touch base, work on problems together, etc. I'm not exactly sure about the boundaries between the -sumo and -develop lists, but my take is that -sumo covers pretty much everything except development topics on specific code (eg, Sigma, CELT). Please forgive (and correct) any errors. Near the end of the course, Adam put up a wish list of areas where he could use some help. I don't have a copy, but here are some general areas that subsume most of his items (and a few that he didn't mention): community building, documentation, etc. software packaging and installation namespace management, syntactic macros, etc. visualization and user experience Some of these areas are already being addressed; others are longer-term and/or highly speculative in nature. In fact, Adam seems quite skeptical about the utility of namespace management, syntactic macros, etc. However, one of the joys of Open Source is the chance to try out partly-baked ideas. Speaking of which, a couple of us are interested in getting Sigma (etc) running under Clojure, then using that environment to start playing with some ideas. Please get in touch if this sounds intriguing... -r -- http://www.cfcl.com/rdm Rich Morin http://www.cfcl.com/rdm/resume rd...@cf... http://www.cfcl.com/rdm/weblog +1 650-873-7841 Software system design, development, and documentation |
|
From: Adam P. <ada...@ea...> - 2007-11-13 22:42:56
|
Hi Meir, Meir wrote: > Adam, > > I would love to acknowledge that the SUMO website is regularly improving. I > did, however, find some entries that need to be reviewed. > > > 1. the third line in: > (=> > (and > (instance ?Card DebitCard) >>>>>>>>> (possesses ?Card ?Agent) <<<<<<<<<<<<< > (instance ?Transaction FinancialTransaction) > (instrument ?Transaction ?Card)) > (exists (?PIN ?Enter) > (and > (pin ?PIN ?Card) > (instance ?Enter EnteringAPin) > (patient ?Enter ?PIN) > (agent ?Enter ?Agent)))) > > should read: > (possesses ?Agent ?Card) > ------------------------------------------------ > fixed > 2. the last line of: > (=> > (instance ?CHART Chart) > (exists (?QUANTITY) > (and > (instance ?QUANTITY PhysicalQuantity) > (refers ?GCHART ?QUANTITY)))) > should be: > (refers ?CHART ?QUANTITY)))) > ------------------------------------------------- fixed > > 3. the following is WRONG. > > (geographicSubregion NitzaneiOzWestBank WestBank) > > Nitzanei Oz was founded in Israel in 1951, so it was never part of the > WestBank. > > (a source in Hebrew, unfortunately. > http://he.wikipedia.org/wiki/%D7%A0%D7%99%D7%A6%D7%A0%D7%99_%D7%A2%D7%95%D7% > 96) > This was just taken in part from data on the Anti-Defamation League web site. I'm reluctant to address correcting the source data, since I'm not an expert in the area, and it's very contentious. > -------------------------------------------------- > 4. The 2 following axioms say the same thing, so one of them may be > redundant. > (=> > (instance ?SENT Sentence) > (exists (?NOUN ?VERB) > (and > (instance ?NOUN NounPhrase) > (instance ?VERB VerbPhrase) > (part ?NOUN ?SENT) > (part ?VERB ?SENT)))) Merge.kif 11699-11706 > > > (=> > (instance ?SENTENCE Sentence) > (exists (?PHRASE1 ?PHRASE2) > (and > (instance ?PHRASE1 NounPhrase) > (instance ?PHRASE2 VerbPhrase) > (part ?PHRASE1 ?SENTENCE) > (part ?PHRASE2 ?SENTENCE)))) Merge.kif 11610-11617 > removed the first one > --------------------------------------------------------- > 5. The similarity between "SportsAttribute" and "SportAttribute" leads me to > believe that only one of them should be in the Ontology. > >> find /n "SportsAttribute" Mid-level-ontology.kif > > ---------- MID-LEVEL-ONTOLOGY.KIF > [12693](subclass SportsAttribute GameAttribute) > [12694](documentation SportsAttribute "Any &%GameAttribute that is specific > to > [13662](instance Foul SportsAttribute) > [13663](documentation Foul "A &%SportsAttribute that indicates that the > &%GamePiece of > [13779](instance SportsOut SportsAttribute) > [13780](documentation SportsOut "A &%SportsAttribute that indicates that a > player in > >> find /n "SportAttribute" Mid-level-ontology.kif > > ---------- MID-LEVEL-ONTOLOGY.KIF > [8367](subclass SportAttribute GameAttribute) > [8368](documentation SportAttribute "Any &%GameAttribute that is specific to > [8374] (instance ?ATTR SportAttribute)) > ---------------------------------------------------------- Changed to SportsAttribute throughout. > 6. I assume this is a typo error (see attached pic1.gif) > Fixed, although since it's minor I'll wait until the next significant upload to SourceForge to release it. The ontology fixes are now on SourceForge but won't be visible in the public browser until I restart Sigma. thanks again! Adam > > Thanks, > > Meir |
|
From: Meir <me...@no...> - 2007-11-13 15:53:53
|
Adam,
I would love to acknowledge that the SUMO website is regularly =
improving. I
did, however, find some entries that need to be reviewed.
1. the third line in:
(=3D>
(and
(instance ?Card DebitCard)
>>>>>>>>(possesses ?Card ?Agent) <<<<<<<<<<<<<
(instance ?Transaction FinancialTransaction)
(instrument ?Transaction ?Card))
(exists (?PIN ?Enter)
(and
(pin ?PIN ?Card)
(instance ?Enter EnteringAPin)
(patient ?Enter ?PIN)
(agent ?Enter ?Agent))))
should read:
(possesses ?Agent ?Card)
------------------------------------------------
2. the last line of:
(=3D>
(instance ?CHART Chart)
(exists (?QUANTITY)
(and
(instance ?QUANTITY PhysicalQuantity)
(refers ?GCHART ?QUANTITY))))
should be:
(refers ?CHART ?QUANTITY))))
-------------------------------------------------
3. the following is WRONG.=20
(geographicSubregion NitzaneiOzWestBank WestBank)
Nitzanei Oz was founded in Israel in 1951, so it was never part of the
WestBank.
(a source in Hebrew, unfortunately.
http://he.wikipedia.org/wiki/%D7%A0%D7%99%D7%A6%D7%A0%D7%99_%D7%A2%D7%95%=
D7%
96)
--------------------------------------------------
4. The 2 following axioms say the same thing, so one of them may be
redundant.=20
(=3D>
(instance ?SENT Sentence)
(exists (?NOUN ?VERB)
(and
(instance ?NOUN NounPhrase)
(instance ?VERB VerbPhrase)
(part ?NOUN ?SENT)
(part ?VERB ?SENT)))) Merge.kif 11699-11706=20
=20
=20
(=3D>
(instance ?SENTENCE Sentence)
(exists (?PHRASE1 ?PHRASE2)
(and
(instance ?PHRASE1 NounPhrase)
(instance ?PHRASE2 VerbPhrase)
(part ?PHRASE1 ?SENTENCE)
(part ?PHRASE2 ?SENTENCE)))) Merge.kif 11610-11617 =20
=20
---------------------------------------------------------
5. The similarity between "SportsAttribute" and "SportAttribute" leads =
me to
believe that only one of them should be in the Ontology.
>find /n "SportsAttribute" Mid-level-ontology.kif
---------- MID-LEVEL-ONTOLOGY.KIF
[12693](subclass SportsAttribute GameAttribute)=20
[12694](documentation SportsAttribute "Any &%GameAttribute that is =
specific
to=20
[13662](instance Foul SportsAttribute)=20
[13663](documentation Foul "A &%SportsAttribute that indicates that the
&%GamePiece of=20
[13779](instance SportsOut SportsAttribute)=20
[13780](documentation SportsOut "A &%SportsAttribute that indicates that =
a
player in
>find /n "SportAttribute" Mid-level-ontology.kif
---------- MID-LEVEL-ONTOLOGY.KIF
[8367](subclass SportAttribute GameAttribute)=20
[8368](documentation SportAttribute "Any &%GameAttribute that is =
specific to
[8374] (instance ?ATTR SportAttribute))
----------------------------------------------------------
6. I assume this is a typo error (see attached pic1.gif)
Thanks,
Meir
|
|
From: <ada...@ea...> - 2007-10-04 15:09:48
|
Folks, Sigma 2.0 has been released at <https://sourceforge.net/project/showfiles.php?group_id=102489> (or from the main project site at http://sigmakee.sourceforge.net). It includes more robust and correct inference performance, and an experimental link to many new theorem provers. It has simplified and more stable installation procedures and many minor enhancements and bug fixes. We'd like your feedback, if you have time to try it out. Adam |
|
From: Adam P. <ada...@ea...> - 2007-07-22 18:35:10
|
Hi Meir, I agree. That's a good addition, and I'll make the change. many thanks, Adam Meir wrote: > Adam, > Your are right. > I would still add "(attribute ?MAMMAL Female)" as it increases the precision > of the statement. > > Meir > >> ---------------------------------------------------------------------- >> --- >> D) The following axiom may imply that if "a bull is a mammal than it gives >> milk". >> >> (=> >> (instance ?MILK Milk) >> (exists (?PROCESS ?MAMMAL) >> (and >> (instance ?MAMMAL Mammal) >> (instrument ?PROCESS ?MAMMAL) >> (result ?PROCESS ?MILK)))) >> >> The correction is to add (attribute ?MAMMAL Female) >> > > Like the reading of the axiom above, this one says "if there is milk, then > there is a mammal that made it" not the other way around. > >> ---------------------------------------------------------------------- > > |
|
From: Meir <me...@no...> - 2007-07-20 15:38:12
|
Adam, Your are right.=20 I would still add "(attribute ?MAMMAL Female)" as it increases the = precision of the statement. Meir > ---------------------------------------------------------------------- > --- > D) The following axiom may imply that if "a bull is a mammal than it = gives > milk".=20 >=20 > (=3D> > (instance ?MILK Milk) > (exists (?PROCESS ?MAMMAL) > (and > (instance ?MAMMAL Mammal) > (instrument ?PROCESS ?MAMMAL) > (result ?PROCESS ?MILK)))) >=20 > The correction is to add (attribute ?MAMMAL Female) > =20 Like the reading of the axiom above, this one says "if there is milk, = then there is a mammal that made it" not the other way around. > ---------------------------------------------------------------------- |
|
From: <ada...@ea...> - 2007-07-20 13:17:14
|
Hi Meir,
Many thanks for the comments. Replies below:
Meir wrote:
> Adam,
>
> Several more questions and comments.
>
> I like the added graphics. Keep up the good work.
>
> Meir
>
>
> -------------------------------------------------------------------
> A) These 2 axioms look contradicting (the 2nd is included in the 1st) (=>
> (and
> (attribute ?AREA FlatTerrain)
> (part ?ZONE ?AREA)
> (slopeGradient ?ZONE ?SLOPE))
> (greaterThan 0.005 ?SLOPE)) Geography.kif 1714-1719
>
> (=>
> (and
> (attribute ?AREA FlatTerrain)
> (part ?ZONE ?AREA)
> (slopeGradient ?ZONE ?SLOPE))
> (greaterThan 0.03 ?SLOPE)) Geography.kif 1726-1731
> -----------------------------------------------------------------------
I've fixed the second axiom to be
(=>
(and
(attribute ?AREA LowTerrain)
(part ?ZONE ?AREA)
(slopeGradient ?ZONE ?SLOPE))
(greaterThan 0.03 ?SLOPE))
> B) I see a conflict between the axiom "(subclass Armor MetalProduct)" and
> the last line in the documentation (quoted from Wikipedia)" to modern
> fabrics such as Kevlar, Dyneema and ceramics."
> "kevlar" is not metallic.
I've changed this to
(subclass Armor Artifact)
> -----------------------------------------------------------------------
> C) the axiom below claims that (my words!)
> "(=>
> (and
> (instance ?A Armor)
> (part ?A ?OBJ))
> (hasPurpose ?A
> (not
> (exists (?DAMAGING)
> (and
> (instance ?DAMAGING Damaging)
> (instrument ?DAMAGING ?A)
> (experiencer ?DAMAGING ?OBJ))))))"
>
> the shell of a turtle and the outer layer of a tank are Armor.
> A ceramic vest that a soldier wears is not part of his body, and, therefore,
> not Armor."
The axiom could be broadened, but I don't believe it's incorrect, or that it conforms to your reading. It says that
"if armor is part of something, then it has the purpose of preventing damage to that thing", not the other way around. However, I've changed it and a related axiom to be:
(=>
(and
(instance ?A Armor)
(or
(part ?A ?OBJ)
(meetsSpatially?A ?OBJ)))
(hasPurpose ?A
(exists (?MAINTAINING)
(and
(instance ?MAINTAINING Maintaining)
(instrument ?MAINTAINING ?A)
(experiencer ?MAINTAINING ?OBJ)))))
(=>
(and
(instance ?A Armor)
(or
(part ?A ?OBJ)
(meetsSpatially?A ?OBJ)))
(hasPurpose ?A
(not
(exists (?DAMAGING)
(and
(instance ?DAMAGING Damaging)
(instrument ?DAMAGING ?A)
(experiencer ?DAMAGING ?OBJ))))))
> -------------------------------------------------------------------------
> D) The following axiom may imply that if "a bull is a mammal than it gives
> milk".
>
> (=>
> (instance ?MILK Milk)
> (exists (?PROCESS ?MAMMAL)
> (and
> (instance ?MAMMAL Mammal)
> (instrument ?PROCESS ?MAMMAL)
> (result ?PROCESS ?MILK))))
>
> The correction is to add (attribute ?MAMMAL Female)
>
Like the reading of the axiom above, this one says "if there is milk, then there is a mammal that made it" not the other way around.
> --------------------------------------------------------------------------
> E) The last line of the "translation" of the right side of (marked with
> "**")
>
> (=>
> (instance ?TACTILE TactilePerception)
> (exists (?TOUCH)
> (and
> (instance ?TOUCH Touching)
> (subProcess ?TOUCH ?TACTILE))))
>
> Merge.kif 10298-10303
> if ?TACTILE is an instance of tactile perception,
> then ?TOUCH is an instance of touching and
> ?TOUCH is a subprocess of ?TACTILE **?TOUCH **
>
> is somewhat confusing. The last ?TOUCH should not appear.
It looks like there's a bug in our NL generation code. We'll look into that.
thanks again!
Adam
>
> ----------------------------------------------------------------------------
> -----
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Sigmakee-sumo mailing list
> Sig...@li...
> https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo
>
|
|
From: Meir <me...@no...> - 2007-07-19 19:31:43
|
Adam,
Several more questions and comments.
I like the added graphics. Keep up the good work.
Meir
-------------------------------------------------------------------
A) These 2 axioms look contradicting (the 2nd is included in the 1st) =
(=3D>
(and
(attribute ?AREA FlatTerrain)
(part ?ZONE ?AREA)
(slopeGradient ?ZONE ?SLOPE))
(greaterThan 0.005 ?SLOPE)) Geography.kif 1714-1719
=20
(=3D>
(and
(attribute ?AREA FlatTerrain)
(part ?ZONE ?AREA)
(slopeGradient ?ZONE ?SLOPE))
(greaterThan 0.03 ?SLOPE)) Geography.kif 1726-1731=20
-----------------------------------------------------------------------
B) I see a conflict between the axiom "(subclass Armor MetalProduct)" =
and
the last line in the documentation (quoted from Wikipedia)" to modern
fabrics such as Kevlar, Dyneema and ceramics."=20
"kevlar" is not metallic.
-----------------------------------------------------------------------
C) the axiom below claims that (my words!)=20
"(=3D>
(and
(instance ?A Armor)
(part ?A ?OBJ))
(hasPurpose ?A
(not
(exists (?DAMAGING)
(and
(instance ?DAMAGING Damaging)
(instrument ?DAMAGING ?A)
(experiencer ?DAMAGING ?OBJ))))))"
the shell of a turtle and the outer layer of a tank are Armor.
A ceramic vest that a soldier wears is not part of his body, and, =
therefore,
not Armor."
-------------------------------------------------------------------------=
D) The following axiom may imply that if "a bull is a mammal than it =
gives
milk".=20
(=3D>
(instance ?MILK Milk)
(exists (?PROCESS ?MAMMAL)
(and
(instance ?MAMMAL Mammal)
(instrument ?PROCESS ?MAMMAL)
(result ?PROCESS ?MILK))))
The correction is to add (attribute ?MAMMAL Female)
=20
-------------------------------------------------------------------------=
-
E) The last line of the "translation" of the right side of (marked with
"**")
(=3D>
(instance ?TACTILE TactilePerception)
(exists (?TOUCH)
(and
(instance ?TOUCH Touching)
(subProcess ?TOUCH ?TACTILE))))=20
Merge.kif 10298-10303=20
if ?TACTILE is an instance of tactile perception,
then ?TOUCH is an instance of touching and=20
?TOUCH is a subprocess of ?TACTILE **?TOUCH **
is somewhat confusing. The last ?TOUCH should not appear.
=20
-------------------------------------------------------------------------=
---
-----
|
|
From: <ada...@ea...> - 2007-07-17 14:12:59
|
Dear Colleague,
Following the success of last year's competition, and the award of the SUMO prize to Robin Sharp of the Technical University of Denmark, we are announcing the second annual SUMO prize for the best open source ontology extension of SUMO.
The SUMO prize is for US$3000.00 will be awarded to the best open source project that extends SUMO. Entries will be due electronically to Adam Pease (apease [at] articulatesoftware [dot] com) by December 1, 2007. Awards will be made December 31, 2007. Entries should be SUO-KIF files that extend SUMO and its domain ontologies, and conform to them.
In addition to the logical soundness of the ontology with respect to the SUMO ontologies, entries will be judged on several criteria:
* Degree of formalization - is the ontology fully specified with rules or just a taxonomy or collection of basic relational statements?
* Scope and coverage of the ontology - is it just a few terms, or hundreds?
* Does the ontology cover a coherent new topic or domain?
* Actual utility of the ontology in an application
Further details are available at <http://www.ontologyportal.org/prize.html>
Please feel free to circulate this notice to your students and colleagues and any mailing lists you feel appropriate.
Adam
|
|
From: Adam P. <ada...@ea...> - 2007-02-05 19:51:42
|
Hi Cameron, As for the graph definitions, it really depends on what you want to do. If you don't need the path start and end in your reasoning application, there's no reason to specify them. A useful set of additions to SUMO would be some mechanism to relate real-world entities that are graph-like to the features of abstract graphs. Part of such additions might be to include default specifications of graph starts and ends. One goal would be to have general axioms about connectivity or reachability defined once for abstract graphs and have those axioms be used if one asks a query about a real-world graph-like entity, such as a road network. Just a thought in case you need a good project :-) Adam Cameron Ross wrote: > Right, procedural programming heritage showing through ;-). Must > BeginNodeFn and EndNodeFn also be explicitly specified for each path > within the graph? Seems like a lot of detail for even moderately sized > graphs. > > Also, is the the proper forum for reporting issues with SUMO? For > example, lines 4326 and 4327 in Merge.kif (pulled from the sigmakee CVS > repository this morning) look incorrect. I believe that: > (links ?ARC1 ?NODE1 ?NODE2) > (links ?ARC2 ?NODE2 ?NODE3) > > should be: > (links ?NODE1 ?NODE2 ?ARC1 ) > (links ?NODE2 ?NODE3 ?ARC2) > > Thanks Adam, > Cameron. > > Adam Pease wrote: >> Hi Cameron, >> This is almost right, just a small syntax issue. Functions still >> employ lisp notation, with the first argument inside the paren. >> >> (equal N1 (InitialNodeFn A12)) >> >> Whenever possible, it's a good idea to load sample code into Sigma, >> since it will catch things like this. >> >> Adam >> >> Cameron Ross wrote: >>> Consider the following directed graph: >>> G1: N1----A12---->N2----A23---->N3 >>> >>> Is the following a proper SUMO representation of G1: >>> (instance G1 DirectedGraph) >>> (instance N1 GraphNode) >>> (graphPart N1 G1) >>> (instance N2 GraphNode) >>> (graphPart N2 G1) >>> (instance N3 GraphNode) >>> (graphPart N3 G1) >>> (instance A12 GraphArc) >>> (graphPart A12 G1) >>> (instance A23 GraphArc) >>> (graphPart A23 G1) >>> (links N1 N2 A12) >>> (equal N1 InitialNodeFn(A12)) >>> (equal N2 TerminalNodeFn(A12)) >>> (links N2 N3 A23) >>> (equal N2 InitialNodeFn(A23)) >>> (equal N3 TerminalNodeFn(A23)) >>> >>> Cheers, >>> Cameron. >>> >>> ------------------------------------------------------------------------- >>> >>> Take Surveys. Earn Cash. Influence the Future of IT >>> Join SourceForge.net's Techsay panel and you'll get the chance to >>> share your >>> opinions on IT & business topics through brief surveys - and earn cash >>> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV >>> >>> _______________________________________________ >>> Sigmakee-sumo mailing list >>> Sig...@li... >>> https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo >>> >> > > |
|
From: Adam P. <ada...@ea...> - 2007-02-05 19:46:03
|
Hi Cameron, Many thanks for the fix. I've posted the new version. Adam Cameron Ross wrote: > Right, procedural programming heritage showing through ;-). Must > BeginNodeFn and EndNodeFn also be explicitly specified for each path > within the graph? Seems like a lot of detail for even moderately sized > graphs. > > Also, is the the proper forum for reporting issues with SUMO? For > example, lines 4326 and 4327 in Merge.kif (pulled from the sigmakee CVS > repository this morning) look incorrect. I believe that: > (links ?ARC1 ?NODE1 ?NODE2) > (links ?ARC2 ?NODE2 ?NODE3) > > should be: > (links ?NODE1 ?NODE2 ?ARC1 ) > (links ?NODE2 ?NODE3 ?ARC2) > > Thanks Adam, > Cameron. > > Adam Pease wrote: >> Hi Cameron, >> This is almost right, just a small syntax issue. Functions still >> employ lisp notation, with the first argument inside the paren. >> >> (equal N1 (InitialNodeFn A12)) >> >> Whenever possible, it's a good idea to load sample code into Sigma, >> since it will catch things like this. >> >> Adam >> >> Cameron Ross wrote: >>> Consider the following directed graph: >>> G1: N1----A12---->N2----A23---->N3 >>> >>> Is the following a proper SUMO representation of G1: >>> (instance G1 DirectedGraph) >>> (instance N1 GraphNode) >>> (graphPart N1 G1) >>> (instance N2 GraphNode) >>> (graphPart N2 G1) >>> (instance N3 GraphNode) >>> (graphPart N3 G1) >>> (instance A12 GraphArc) >>> (graphPart A12 G1) >>> (instance A23 GraphArc) >>> (graphPart A23 G1) >>> (links N1 N2 A12) >>> (equal N1 InitialNodeFn(A12)) >>> (equal N2 TerminalNodeFn(A12)) >>> (links N2 N3 A23) >>> (equal N2 InitialNodeFn(A23)) >>> (equal N3 TerminalNodeFn(A23)) >>> >>> Cheers, >>> Cameron. >>> >>> ------------------------------------------------------------------------- >>> >>> Take Surveys. Earn Cash. Influence the Future of IT >>> Join SourceForge.net's Techsay panel and you'll get the chance to >>> share your >>> opinions on IT & business topics through brief surveys - and earn cash >>> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV >>> >>> _______________________________________________ >>> Sigmakee-sumo mailing list >>> Sig...@li... >>> https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo >>> >> > > |
|
From: Cameron R. <cr...@sy...> - 2007-02-05 14:39:57
|
Right, procedural programming heritage showing through ;-). Must
BeginNodeFn and EndNodeFn also be explicitly specified for each path
within the graph? Seems like a lot of detail for even moderately sized
graphs.
Also, is the the proper forum for reporting issues with SUMO? For
example, lines 4326 and 4327 in Merge.kif (pulled from the sigmakee CVS
repository this morning) look incorrect. I believe that:
(links ?ARC1 ?NODE1 ?NODE2)
(links ?ARC2 ?NODE2 ?NODE3)
should be:
(links ?NODE1 ?NODE2 ?ARC1 )
(links ?NODE2 ?NODE3 ?ARC2)
Thanks Adam,
Cameron.
Adam Pease wrote:
> Hi Cameron,
> This is almost right, just a small syntax issue. Functions still
> employ lisp notation, with the first argument inside the paren.
>
> (equal N1 (InitialNodeFn A12))
>
> Whenever possible, it's a good idea to load sample code into Sigma,
> since it will catch things like this.
>
> Adam
>
> Cameron Ross wrote:
>> Consider the following directed graph:
>> G1: N1----A12---->N2----A23---->N3
>>
>> Is the following a proper SUMO representation of G1:
>> (instance G1 DirectedGraph)
>> (instance N1 GraphNode)
>> (graphPart N1 G1)
>> (instance N2 GraphNode)
>> (graphPart N2 G1)
>> (instance N3 GraphNode)
>> (graphPart N3 G1)
>> (instance A12 GraphArc)
>> (graphPart A12 G1)
>> (instance A23 GraphArc)
>> (graphPart A23 G1)
>> (links N1 N2 A12)
>> (equal N1 InitialNodeFn(A12))
>> (equal N2 TerminalNodeFn(A12))
>> (links N2 N3 A23)
>> (equal N2 InitialNodeFn(A23))
>> (equal N3 TerminalNodeFn(A23))
>>
>> Cheers,
>> Cameron.
>>
>> -------------------------------------------------------------------------
>>
>> Take Surveys. Earn Cash. Influence the Future of IT
>> Join SourceForge.net's Techsay panel and you'll get the chance to
>> share your
>> opinions on IT & business topics through brief surveys - and earn cash
>> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>>
>> _______________________________________________
>> Sigmakee-sumo mailing list
>> Sig...@li...
>> https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo
>>
>
|
|
From: Adam P. <ada...@ea...> - 2007-02-05 01:02:15
|
Hi Cameron, This is almost right, just a small syntax issue. Functions still employ lisp notation, with the first argument inside the paren. (equal N1 (InitialNodeFn A12)) Whenever possible, it's a good idea to load sample code into Sigma, since it will catch things like this. Adam Cameron Ross wrote: > Consider the following directed graph: > G1: N1----A12---->N2----A23---->N3 > > Is the following a proper SUMO representation of G1: > (instance G1 DirectedGraph) > (instance N1 GraphNode) > (graphPart N1 G1) > (instance N2 GraphNode) > (graphPart N2 G1) > (instance N3 GraphNode) > (graphPart N3 G1) > (instance A12 GraphArc) > (graphPart A12 G1) > (instance A23 GraphArc) > (graphPart A23 G1) > (links N1 N2 A12) > (equal N1 InitialNodeFn(A12)) > (equal N2 TerminalNodeFn(A12)) > (links N2 N3 A23) > (equal N2 InitialNodeFn(A23)) > (equal N3 TerminalNodeFn(A23)) > > Cheers, > Cameron. > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys - and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Sigmakee-sumo mailing list > Sig...@li... > https://lists.sourceforge.net/lists/listinfo/sigmakee-sumo > |
|
From: Cameron R. <cr...@sy...> - 2007-01-25 00:25:28
|
Consider the following directed graph: G1: N1----A12---->N2----A23---->N3 Is the following a proper SUMO representation of G1: (instance G1 DirectedGraph) (instance N1 GraphNode) (graphPart N1 G1) (instance N2 GraphNode) (graphPart N2 G1) (instance N3 GraphNode) (graphPart N3 G1) (instance A12 GraphArc) (graphPart A12 G1) (instance A23 GraphArc) (graphPart A23 G1) (links N1 N2 A12) (equal N1 InitialNodeFn(A12)) (equal N2 TerminalNodeFn(A12)) (links N2 N3 A23) (equal N2 InitialNodeFn(A23)) (equal N3 TerminalNodeFn(A23)) Cheers, Cameron. |
