6th International Conference on Web Information Systems Engineering (WISE 2005)
21-22 November 2005, New York City
6th International Conference on Web Information Systems Engineering (WISE 2005)
New York City, Intercontinental Barclay Hotel
Author: Raphael
CWI participants: Raphael
# participants: 100
Overall impression
About 100 participants to this 6th edition of the WISE conference, less than usually. WISE is still a high selective conference (260 papers
received for about 12% accepted as long papers) which covers a very broad range of topics. The two days conference were split into 3
parallel sessions: Web mining, Web information retrieval, Metadata management, Ontology and Semantic Web, XML, Web service method,
Web service structure, Collaborative methodology, P2P Ubiquitous and Mobile and Document retrieval applications. I detail below only my
session, but the full proceedings is available in my office.
Research Session 4: Ontology and Semantic Web
13:30 - A Formal Ontology Reasoning with Individual Optimization: A Realization of the Semantic Web
Pakornpong Pothipruk, Guido Governatori
Funny introduction to Description Logics (it was good since I hadn't have to present what is OWL in my talk afterwards :-). Problem: in the
context of the Web, the ABox will be very large, thus problem of efficiency of query answering. Two standard queries in DL: ABox instances
checking and ABox instances retrieval. Focus on the second type of queries (i.e. non-boolean queries).
Solution: space partitioning and reduction approach. The instances are splitted into several KB. The paper presents the algorithm to make
such a partition:
- Find the dependencies among the ABoxes
- Group the non dependent ABoxes
- Query indepently each group and merge the results
The paper defines formally what is an independent ABox. The strong asumption behind is that you can find these independent ABoxes !
Furthermore, the audience was not really convinced that this kind of optimization is not currently available in system like Racer. I'm
personnaly not convinced that this trick will allow to decrease of a complexity class problem (i.e. go down from NExpTime to ExpTime or
Polynomial/Tractable)!
14h00 - oMAP: Combining Classifiers for Aligning Automatically OWL Ontologies
Umberto Straccia, Raphael Troncy
I think the audience has mainly understood my talk. The questions were:
- How can we be sure that the alignments are good: because in the benchmark tests, the good ones were assessed by human hand and checked by
a committee;
- Aligning ontologies, for doing what? A quite long discussion has followed, and I'm happy about that since it is basically the future work
I plan to do. Obviously, aligning ontologies is not an objective per su, and it should be integrated into a larger scenario. One possible
realistic scenario is to make some query reformulation: the idea is, given a query formulated according to a KB1, be able to reformulate
this query according to a KB2 via the alignment of their model, and merge the results. It will be the main topic of our next paper for ESWC,
"Towards Distributed Search on the Semantic Web".
14h30 - Semantic Web Technologies for Interpreting DNA Microarray Analyses: The MEAT System
Khaled Khelif, Rose Dieng-Kuntz, Pascal Barbry
My Colleagues from INRIA Nice who present their ideas on corporate Semantic Web. Description of the Meat ontology (a biological ontology)
and demonstration of the Corese reasoner, which consumes RDFS ontology, RDF data and implements almost all the SPARQL query
language.