Scribe log for 13 December 2001 teleconference meeting

		Scribe Log for WebOnt 13 December 2001

CHAIR: Dan Connolly
SCRIBE: Peter F. Patel-Schneider

> 1) Administrivia  (10 min, includes telecon greets and attendance)

Roll call (by DanC)

New Agenda Item: Peter and Frank - subgroup 

> ACTION (Decker, Schreiber, Obrst, Heflin, McGuinness):  Within the 
> next week, produce outline of area of use cases.>
>    status: to be discussed, agenda item 3

See below

> ACTION (JimH): solicit/appoint editors/process for producing 
> completed document from these pieces
>   status: area leads identified, process emerging - continue action

Continued - need to find out what Jim actually means by this

> 2) DAML+OIL Q&A (Ian Horrocks, Frank van Harmelen) 20 min
>   Chance for WG members to ask questions about DAML+OIL - walkthru or beyond.

Frank van Harmelen is available for questions
Dan Connolly asks for questions

Leo Obrst: What is the status of rules?
Frank: Useful, necessary, but explicitly out of scope

??:  What about quantification?
Frank:  Limited because of computational issues.

??: What about equality?
Frank:  Included, no unique name assumption.

??: What about things like uncle?
Frank:  Can't be done.  

DanC: Comments on role rdf-logic, rdf-rules, ... mailing lists
      Role of rules.

Leo Obrst: What about metalevel facilities?
Frank: Status is unclear.  How would it work?
Jeremy Caroll: Can you use subProperty/subClass to get some of this?
Frank: Sort of.
Jeremy Caroll: What about annotation.  E.g., can you require annotations on
Frank:  Yes, but you don't get any semantic extensions.
Peter: No change to behaviour will arise.
Leo: What about reification?
DanC cuts this off

Harold?: Can you chain properties?
Frank: Not in DAML+OIL.
Peter: Decidability issues arise here.

4) New Business

- Cleaning up DAML+OIL - Peter
- Revising DAML+OIL - Frank - actually feedback on problems

Peter has volunteers - six other people
- goal is to get documents ready for the f2f

Dieter +1 : all to group
Lynn : wants to have pointers to all relevant messages
? : categorize mail messages on webont

Peter will pick a mode for email message for his effort

Proposal: let groups determine their email mode

ACTION:  Peter will produce document on cleanup by 7 January

DanC: what about Frank's task?

Frank doesn't have enough time to do it

Mike Dean volunteers to collect and gather experience reports

ACTION: Mike Dean will produce document on experience with DAML+OIL (and
	suggestions for revisions?) by 7 January (non-negotiable)

> 3) Use case document[s] - 45 min
>    Discussion led by chairs of the area - each will discuss status and 
> outline, open discussion thereafter.

** Stefan Decker: Web Services - 
   - email sent 13 December - use-case area summary: Web Services
- interesting portions - combining web services
		       - searching for web services
- use case - device interconnection and interoperation

DanC - chair message - use cases should not be ``exhaustive'' as then we
		       will end up with an undoable task
     - instead try for a typical, important use case
     - group will then determine which use cases must be supported
       - need to have consensus to focus the work
??? - also need to be accessible to and compelling for outsiders
DanC - also need to be concrete

DanC - Q: status of Stefan's portion?

Stefan Decker - status - less than 1/3 done
		       - no specific examples
Jeremy Caroll - sample implementation(?) for device interconnection
	      - will send in if possible

??: use cases should be important 
Deb: agree - use cases should not be picked just to show off DAML+OIL

** Guus Schreiber: Collection management
   - email sent 13 December - use-case area summary: collection management
- small number of use cases
- Arkive use case - endangered species descriptions
  - important ideas - default inheritance

Stephen Buswell - engineering data modelling is typical use case
	- can this be done in DAML+OIL

??? - want concrete examples for use case

Guus - status of portion - need concrete example, e.g., from Arkive

Guus - what format to use?

Frank - think of implications for language design from use cases
Guus - there are some in there
Frank - actually a call for (all) other groups to do the same

** Leo Obrst
   - email sent 13 December - Use Case: Content Interoperability

- what does content interoperability cover - device example above??

DanC - don't need to define content interoperability exactly, just get use
     cases that *are* in it

- basic goals - ..., mappings of ontologies (e.g., catalogues)
- features needed - ....

- travel scenario use case (from Mike Dean)

DanC - a use case is a specific scenario
     [pfps] i.e., something that could be directly implemented

- status - 1/3 done

- contact with upper ontology group

** Jeff Heflin - General Requirements
   - email sent 13 December - General Requirements Subgroup - Progress Report

- email discussion resulted in 16 general requirements
- six have broad support - shared ontologies
			 - ontology extension
			 - scalability
			 - ease of use
			 - XML syntax
			 - expressiveness
- others are under question - unclear, out of scope, too hard

- next steps - finalize status
	     - detailed requirements for chosen few

Mike Dean - provenance of facts
Jeff - general idea of tagging has come up

- status - 1/3 + ``a little'' done - less than 1/2

DanC - internationalization?
Jeff - character part from XML
     - different terms from interoperability
DanC - political considerations make internationalization important
Lynn - use better terms instead of ``internationalization''

various - multlingual support is non-optional

DanC - next meeting?
     - next week - 20th

Scribe for next week not yet determine

Agenda - cleanup committee status
       - ...

Meeting closed

Post-closing notes (from pfps)

On reading the above it appears that ACTION (JimH) was to get someone to
pull together the four use-case documents into one.  Something probably
needs to be done on this very soon.

Received on Thursday, 13 December 2001 13:41:09 UTC