* Faculty       * Staff       * Students & Alumni       * Committees       * Contact       * Institute Directory
* Undergraduate Program       * Graduate Program       * Courses       * Institute Catalog      
* Undergraduate       * Graduate       * Institute Admissions: Undergraduate | Graduate      
* Colloquia       * Seminars       * News       * Events       * Institute Events      
* Overview       * Lab Manual       * Institute Computing      
No Menu Selected

* News


Web Service Composition as Planning: A New Perspective

Joerg Hoffmann
University of Innsbruck, Semantic Technologies Institute, Innsbruck, Austria

Monday, January 7, 2008
JEC 3117 - 11:00 a.m. to 12:00 a.m.
Refreshments at 10:30 a.m.


The talk consists of three parts, the first of which gives a brief glimpse of my research background concerning the FF Planning tool and its performance, and the last of which overviews my current lines of ongoing research. The middle part is a technical talk on a particular topic of current research, concerning the automated composition of Semantic Web services described at the functional level, in terms of logical preconditions and effects. The main contribution of this work is a novel compilation of the composition problem into Planning, enabling us to exploit the power of FF to solve WSC tasks involving complex ontologies. Starting from a formalization of WSC, I explain how computing the outcome of Web service applications corresponds to what AI calls "belief update", and how this affects the associated computational complexity. I then identify a special case, termed forward effects, where Web services mainly affect their outputs; this special case corresponds to what is often called "message-based composition". I show that belief update simplifies in that special case, and I show how the reduced complexity enables us to define a compilation into Planning under initial state uncertainty. Reasoning about uncertainty in the compiled task is the same as reasoning about the properties of outputs in the original task. I demonstrate that, this way, large and complex composition tasks can be solved using FF off-the-shelf.

More information:

Dr. Hoffmann's webpage.

Hosted by:

James Hendler (x4401) and Deborah McGuinness (x4404)
Administrative support: Jacqueline Carley (x4384)

Last updated: December 29, 2007