Joint SIO-OOR-Ontolog Panel Discussion: "CL Support for OOR" - Fri 10-Sep-2010    (2GTA)

Conference Call Details:    (2GTR)

Attendees:    (2GUQ)

Resources    (2GV3)

Agenda & Proceedings:    (2GVH)

Session: "Common Logic (CL) Support for OOR" Panel Discussion    (2GVI)

Abstracts:    (2GVO)

Abstract: ... this session will be devoted to exploring the various alternatives available to provide Common Logic (ISO/IEC 24707) support for the Open Ontology Repository (OOR) initiative.    (2GVQ)

Panel Member Talks:    (2GVR)

Abstract: ... The Heterogeneous Tool Set (Hets) is a multi-logic parsing, analysis and proof tool. The logics covered by Hets include propositional logic (with SAT solvers minisat and zChaff), OWL (with provers Pellet and FACT++), first-order logic (with provers SPASS, Vampire, Darwin and MathServe), and higher-order logic (with prover Isabelle). Recently, we have added Common Logic to Hets, offering both parsing and proof tools for Common Logic. SPASS, Vampire, Darwin and MathServe are available for proofs in Common Logic, as well as Isabelle (if induction on lists is needed). Moreover, a translation from OWL into Common Logic is provided. Last but not least, the language HetCASL provides heterogeneous structuring and refinement, making it possible e.g. to verify that the OWL version of an ontology is indeed implied by its Common Logic version, or verify logical relations between different ontologies. We propose functionalities of this kind as highly desirable for next generation ontology management and repository building.    (2HOD)

Abstract: I will give an overview of the objectives for the COLORE project, covering ontological engineering, issues specific to Common Logic, and issues related to software functionality.    (2GVX)

Abstract: ... This talk will outline efforts by Kojeware Corporation and the Semantic Technologies Laboratory at the University of Toronto to develop an open system for hosting a repository of Common Logic content from Goolge's cloud computing infrastructure. The requirements of this system will be summarized and a high-level description of the system will be presented. The Resource-Oriented Architecture we have designed to expose CL content using RESTful services will also be presented. Challenges with naming CL resources using globally unique and location-independent identifiers (i.e. URNs) will be discussed, and disparities between our requirements and capabilities of the current BioPortal system will be summarized.    (2GW3)

ref. Announcement - Open Source release of Samian Platform:    (2GWU)

Abstract: ... This brief talk will outline the capabilities of the open source (under GNU GPL/LGPL) Sigma Knowledge Engineering Environment Sigma has been released and developed on SourceForge starting in 2004. This talk will discuss its existing tool set including browsing, analysis and inference. Sigma supports SUO-KIF, with import/export capabilities for OWL, and TPTP formats. Its embedded inference system supports full first order logic with equality. An experimental higher order inference system is also included. While developed for the Suggested Upper Merged Ontology, Sigma can support any theory or set of theories of equivalent or lesser expressiveness. The SUO-KIF appears to be a concrete syntax of CL, and little effort would be required to allow CLIF display/import/export within Sigma.    (2GW0)

Transcript of the online chat during the session:    (2GW7)

 see raw transcript here.    (2GW8)
 (for better clarity, the version below is a re-organized and lightly edited chat-transcript.)
 Participants are welcome to make light edits to their own contributions as they see fit.    (2GW9)
    -- begin of chat session --    (2GWA)
	PeterYim: .    (2HTT)
	Welcome to the Joint SIO-OOR-Ontolog Panel Discussion - Fri 10-Sep-2010    (2HTU)
	* Topic: "Common Logic (CL) Support for OOR"    (2HTV)
	* Chair: Dr. JohnSowa (VivoMind Intelligence)    (2HTW)
	* Panelists:
	  o Professor JohnBateman and Dr. habil. TillMossakowski (Bremen U) 
	     - "Adding Common Logic Support to the Heterogeneous Toolset (status report)"
	  o Professor MichaelGruninger (U of Toronto) 
	     - "Objectives of the Common Logic Ontology Repository Project"
	  o Mr. CameronRoss (Kojeware) 
	     - "An OOR Implementation for COLORE"
	  o Mr. AdamPease (Articulate Software) 
	     - "Sigma as an OOR"
	  o Dr. PatHayes - some comments on work presented at this session    (2HTX)
	please refer to dial-in and other details (incl. slide download links) on the session page
	at:    (2HTY)
	.    (2HTZ)
	Please change your name from 'anonymous' using the Settings button (top center)    (2HU0)
	anonymous morphed into vnc2    (2HU1)
	anonymous morphed into TomRuss    (2HU2)
	anonymous morphed into JohnBateman    (2HU3)
	anonymous morphed into MattHettinger    (2HU4)
	DougFoxvog: put the video link on the chat page    (2HU5)
	JohnBateman: The Hets demo video is at:    (2HU6)
	anonymous morphed into ElizabethFlorescu    (2HU7)
	PeterYim: = JohnSowa presenting (started 9:13am PDT) ...    (2HU8)
	DougFoxvog: Can the Prolog cut operator be easily handled by First-Order logic?    (2HU9)
	AliHashemi: i think it should be on slide in the vnc    (2HUA)
	AliHashemi: slide 5*    (2HUB)
	DougFoxvog: The slide should be advance to slide 5 as Ali said    (2HUC)
	DougFoxvog: Slide 7 now!    (2HUD)
	TillMossakowski: A smaller version (3.5GB) of the Hets demo video is at:    (2HUE)
	AliHashemi: (the link to JohnB's pdf version is not working)    (2HUF)
	AliHashemi: try VLC ?    (2HUG)
	AliHashemi: otherwise you might need to update some codecs    (2HUH)
	DougFoxvog: I can play it with the VLC media player.    (2HUI)
	RickMurphy: For folks who are interested. Hets is easy to install and config. Need to lean a little 
	Haskell, understand the background and you're good to go.    (2HUJ)
	TillMossakowski: However, you can install and run Hets without any Haskell knowledge. Haskell 
	knowledge is only needed if you want to develop Hets.    (2HUK)
	TomRuss: Quicktime on Mac is unable to play the movie.    (2HUL)
	CameronRoss: I'm uploading the small video to youtube... it should be available shortly    (2HUM)
	RickMurphy: Till's right. I was just think about those who want to get their hands dirty. Also great 
	support by Christian Maeder !    (2HUN)
	CameronRoss: try and set your player to 480p    (2HUO)
	PeterYim: thanks, Cameron ... this would work (probably for most of us)    (2HUP)
	CameronRoss: Just click the link above and it will start in your browser    (2HUQ)
	anonymous2 morphed into PavithraKenjige    (2HUR)
	CameronRoss: @Peter - ctrl L will put Adobe Reader into fullscreen    (2HUS)
	PeterYim: the youtube link is now posted to the session page too    (2HUT)
	PeterYim: @Cameron - I am aware of that, but when the slide decks don't show a slide number, I'll 
	need to stay on the smaller window to know what slide number we are on    (2HUU)
	CameronRoss: @Peter - makes sense    (2HUV)
	PeterYim: = MichaelGruninger presenting (start 9:59am PDT) ...    (2HUW)
	ToddSchneider: Thank you to all presenters. Have to go.    (2HUX)
	PeterYim: bye, Todd    (2HUY)
	PeterYim: = CameronRoss presenting (started 10:09am PDT) ...    (2HUZ)
	anonymous2 morphed into PatHayes    (2HV0)
	PatHayes: hi all, PatHayes just got online.    (2HV1)
	PeterYim: Hi Pat ... glad you made it!    (2HV2)
	PatHayes: hi peter. which slides are we in?    (2HV3)
	PeterYim: please morph into your real name (use "Settings" button at top center of window)    (2HV4)
	PeterYim: CameronRoss's slide#4    (2HV5)
	JohnBateman: The high resolution version of the HETS/CL demo video is now also on youtube at    (2HV6)
	PeterYim: Thanks John & Till ... links on session pages updated - ref    (2HV7)
	AdamPease: phone battery died dialing back in    (2HV8)
	JohnBateman: There seems to be a terminological confusion between "Common Logic" as the CL that 
	appears in CLIF files, and "Common Logic" in the broader sense. Otherwise it does not really make 
	sense to talk of a "CL dedicated reasoner": what is necessary surely is the connection between many 
	different reasoners as appropriate for the expressivity of the particular sublogics that CL may 
	define (as argued for by JohnSowa just now): i.e., what Hets does.    (2HV9)
	PatHayes: johnB: yes, people often say CL when they really mean CLIF. However, there is what one 
	might call a natural bias in CL to cover the whole free cl syntax, which goes beyond KIF in 
	significant ways.    (2HVA)
	anonymous6 morphed into BartGajderowicz    (2HVB)
	anonymous6 morphed into TillMossakowski    (2HVC)
	CameronRoss: @Adam - One issue I've found with translating SuoKif to CL is that SuoKif allows terms 
	to be sentences. However, over the thousands of sentences contained within SUMO, there are only 57 
	sentences that use this construct.    (2HVD)
	TillMossakowski: We have developed a technique for proving consistency of ontologies by amalgamating 
	small models, conservative extensions, and refinements. We have applied this to DOLCE, which is not 
	as huge as SUMO, but larger then anything model finders can handle.    (2HVE)
	PeterYim: I was disconnected from this chat-session (the server may have been down) just 
	now ... and have lost the chat-transcript prior to now ... if anyone still has that, kindly say so 
	(and send me a copy for the archives, please    (2HVF)
	BartGajderowicz: @PeterYim: I have it, will archive and forward to you, Peter    (2HVG)
	PeterYim: thanks, Bart!    (2HVH)
	JohnBateman: Note to Adam on CVS: actually there are many reasons to go beyond basic text CVS. We 
	are currently exploring the use of the XML-aware versioning system TnTBase (probably to be talked 
	about next week) and interfacing this with BioPortal. There is also work ongoing on semantic object 
	versioning, which is where we will want to be for ontology engineering I expect. TnTBase is, 
	however, built *on top* of standand versioning tools, so I would agree with Adam that there should 
	be no duplication of effort or repeating work that already exists, just because we are looking at 
	ontologies.    (2HVI)
	anonymous2 morphed into RickMurphy    (2HVJ)
	anonymous9 morphed into MichaelGruninger    (2HVK)
	anonymous morphed into ImmanuelNormann    (2HVL)
	JohnBateman: For alignment and translation between ontologies, we have moved on somewhat: I'd refer 
	you to the work from TillMossakowski, Oliver Kutz, Immanuel Normann on 'shapes of alignment'. This 
	makes much of the discussion about fixing particular vocabularies redundant.    (2HVM)
	DougFoxvog: The 90% ^ 5 => 6% argument only holds if each step is independent of the previous ones. 
	However, if one is looking at logic chains, the separate steps are likely to be in the same context. 
	Thus if the 90% covers the first step, the likelihood that terms in following rules also correspond 
	should be far higher than 90%.    (2HVN)
	anonymous morphed into JohnSowa    (2HVO)
	AdamPease: I'm tempted to comment on the ontology discussion, but isn't it very off topic? This call 
	is about an OOR, not upper ontology    (2HVP)
	JohnBateman: I agree with Adam: old debates.    (2HVQ)
	AliHashemi: Depending on how the modules are related to one another (Conservative vs Non 
	Conservative) it goes a long way in addressing PatCassidy's point    (2HVR)
	CameronRoss: @AdamPease and @JohnBateman - I'm currently looking into the possibility of integrating 
	and re-branding the Google Code hosting service for an OOR.    (2HVS)
	JohnBateman: Pat: before you go.... can we have agreement on the module semantics!!! We're waiting 
	on this to finalise the Hets treatment of the modules.    (2HVT)
	CameronRoss: @PatHayes - I have a pile of comments regarding a revised spec! I'll start a dialog.    (2HVU)
	anonymous1: Where's the CL mailing list?    (2HVV)
	PeterYim: who is this? "anonymous1"    (2HVW)
	AliHashemi:    (2HVX)
	JohnBateman: @PatHayes: we *need* the module decision!!!    (2HVY)
	CameronRoss: @PatCassidy - in terms of translation, the Samian Platform will support arbitrarily 
	complex transformations steps, including steps that require complex reasoning... just need to know 
	what thot is    (2HVZ)
	PavithraKenjige: @Dr.Sowa I am not in the common logic group, but I was interested in listen ingto 
	this group Thanks! Thanks every body!    (2HW0)
	PeterYim: thanks everyone ... great session!    (2HW1)
	AliHashemi: Thanks + Bye!    (2HW2)
	CameronRoss: Bye everyone... thanks!    (2HW3)
	PeterYim: -- session ended 11:10am PDT --    (2HW4)
	TillMossakowski: good bye    (2HW5)
	JohnBateman: See y'all.    (2HW6)
	PeterYim: @Bart ... looks like I do have a copy of the chat-transcript after all (from the vnc 
	server) ... therefore, no need to send, but really appreciate your offer to help, nonetheless!    (2HW7)
	BartGajderowicz: @PeterYim, no problem    (2HW8)
	PeterYim: Thanks again, Bart ... ttyl!    (2HW9)
    -- end of chat session --    (2GWB)

Audio Recording of this Session    (2GWF)

For the record ...    (2GWN)

How To Join (while the session is in progress)    (2GWO)