chat-transcript_20100401b_unedited.txt


PeterYim: .

Welcome to the Joint OOR-Ontolog Panel Discussion: "OOR Use Cases - Take-3" - Thu 01-Apr-2010

* Co-chairs: Professor KenBaclawski and Dr. ToddSchneider    (2D15)

* Panelists: 
  o Dr. ToddSchneider (Raytheon)
  o Mr. MatthewHettinger (Mathet Consulting)
  o Professor KenBaclawski & Mr. MaximoGurmendez (Northeastern U) 

please refer to the session page for dial-in, agenda, slides, etc.
at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2010_04_01 

.

anonymous morphed into Frank Hanley

anonymous morphed into MattHettinger

TerryLongstreth: @Todd: what means 'not hierarchical'?

PeterYim: @Todd: why are you suggesting that "not a near-realtime system" ... should it rather be "some services may be be realtime or near-realtime"?

PeterYim: above should read: why are you suggesting that the OOR is "not a near-realtime system" ...

anonymous morphed into Bobbin Teegarden

TerryLongstreth: @Todd: use case for Role: Inference engine?

PeterYim: @Todd: should we disassociate the Ontology Repository from the Ontology Editor (or the ontology design and development platform)?

DavidEddy2: Peter: "not near-realtime" because the number of thingees/connections tend to be HUGE

CameronRoss: @Todd: what about the provenance of IP?  Should contributions to the repository be limited to those that have signed some kind of an IP release?  This is often the case for large open source projects such as Eclipse etc.

CameronRoss: soft-real time

CameronRoss: @Peter:  I would hope that the OOR would provide some kind of service-oriented interface for a variety og Ontology Development Environments.  An ODE will most likely be optimized for a particular representation language.  Therefore, decoupling in this way would help promote support for multiple representations.

PeterYim: @Cameron: agree totally

DavidEddy2: translation... SIO = ?

PeterYim: @David: SIO = the newly established "Sharing and Integrating Ontologies (SIO)" project - see: http://ontolog.cim3.net/forum/ontolog-forum/2010-04/msg00000.html

TerryLongstreth: In general, I believe it will be hard to separate the individual ontologies in the repository. I assume that for most purposes any use of the OOR content may involve access (or execution?) of more than one contributed module. I prefer to think of the repository as holding an evolving, single ontology, with varying degrees of consistency among its constituents, and varying degrees of authenticity or correctness within any one constituent.

TerryLongstreth: It's hard to know whether this perspective would be addressed in OOR or SIO requirements.

alex: What about the evaluation u mentioned in the beginning? slide 3

DavidEddy2: Are humans going to look into the OOR for something useful?   Or is it only software programs that will hit on the OOR?

alex: I think onto designers will look into the OOR via their ontology editors.

PeterYim: @David: should be BOTH humans and machines

alex: should be both yes

DavidEddy2: what is distinction between "version management" and "configuration management"?

MattHettinger: @alex e.g. does the use of a particular ontology or group of ontologies increase business performance by decreasing communication errors in human-human, human-machine, machine-machine interactions. One may create a Design of Experiments to test the hypothesis. Factor entering into acceptance or rejection will depend on the quality of ontology, the appropriate use of the ontology, the underlying assumptioins, etc. all of which could be parsed out in the evaluation. This is one example. Does this help?

alex: Matt, I get it

PeterYim: @Ken: given that the bottleneck from our moving from the OOR-sandbox to an OOR production instance is the gatekeeping mechanism and the associated policies that a production OOR need ... what do you see as the immediate next steps and the estimated time frame before we can see our first production OOR (instance)?

alex: fit to purpose, being the purpose that of the business. am I right?

alex: @Matt fit to purpose, being the purpose that of the business. am I right?

MattHettinger: @alex yes, though each subsystem may have its own purpose, they need to be aligned w/ the business as a whole

TerryLongstreth: I have to sign off.  I'll look back at the wiki tomorrow.

alex: @ all, Maximo is right about licensing.

PeterYim: (while we have yet to finalize and document IPR policies for OOR) I need to remind contributors that the general direction we are going is "open source licensing" for the technology, and "creative commons licensing" for the content (i.e. ontologies) ... and also that OOR should be designated and implemented to federate with other non-open ontology repositories

PeterYim: correction: above should read "designed and implemented ... "

PeterYim: great session!

PeterYim: thank you for your participation and contributions

PeterYim: -- session ended: 2010.04.01 12:29pm PDT --