RE: [uima] Re: [CAS Specification Subgroup] Getting Started

[David Ferrucci <ferrucci@us.ibm.com>]

Our intention was  to allow the
freedom provided by the Regional Reference but to include "standard"
annotation types in the base-type system to address interoperability issues.
I think its important to define certain
base types exactly to encourage/support interoperability. At the same time
I do not think we want to get into the business of defining base annotation
types for every possible artifact representation. So I have argued we need
both.


------------------------------------------------------------------------
David A. Ferrucci, PhD
Senior Manager, Semantic Analysis & Integration 
Chief Architect,  UIMA
IBM T.J. Watson Research Center
19 Skyline Drive, Hawthorne, NY 10532
Tel: 914-784-7847, 8/863-7847
ferrucci@us.ibm.com
------------------------------------------------------------------------
http://www.ibm.com/research/uima

"Scott Songlin Piao" <scott.piao@manchester.ac.uk>

03/01/2007 07:06 AM

To	"Eric Nyberg" <ehn+@cs.cmu.edu>, "Sophia.ananiadou@manchester.ac.uk" <Sophia.ananiadou@manchester.ac.uk>, "tgoetz@de.ibm.com" <tgoetz@de.ibm.com>, "ehn@cs.cmu.edu" <ehn@cs.cmu.edu>, Ngan <nltngan@is.s.u-tokyo.ac.jp>, "carl.madson@sri.com" <carl.madson@sri.com>, "kano@is.s.u-tokyo.ac.jp" <kano@is.s.u-tokyo.ac.jp>, "j.tsujii@manchester.ac.uk" <j.tsujii@manchester.ac.uk>
cc	"uima@lists.oasis-open.org" <uima@lists.oasis-open.org>
Subject	RE: [uima] Re: [CAS Specification Subgroup] Getting Started

Hi,

Here is my comment on the regional reference of Annotation type (section
5.1).

1) According to Fig. 3 (Sofa Reference and Regional Reference), the Annotation
type itself does not have regional reference features, it only has sofa
reference. The paragraph above Fig. 3 says that any subtype of the Annotation
MAY define how the regional reference is implemented. Does it imply that,
by definition, the Annotation is not necessarily linked to certain region
of the data? This seems to conflict against the general definition of Annotation
as a type linked to certain region.

2) By allowing the freedom for subtypes of Annotation to define the way
the regional reference is implemented, it may result in specialized implementations
of the regional reference, which may potentially cause problem for interoperability.

3) As have been mentioned in previous discussions already, the multi-region
reference is an issue to be resolved.

Best regards

Scott Piao