Uncertainty Handling in the Context of Ontology Mapping for Question-Answering Miklos Nagy
This event took place on 17th January 2007 at 11:30am (11:30 GMT)
Knowledge Media Institute, Berrill Building, The Open University, Milton Keynes, United Kingdom, MK7 6AA
The combination of different similarity methods in the current ontology mapping approaches can considerably increase the quality of the mappings however uncertainty caused by incomplete or inconsistent data has received relatively little attention in the ontology mapping community. This paper describes a framework for integrating similarity measures and Dempster-Shafer belief functions for ontology mapping in the context of multi agent ontology mapping. Our novel approach describes how to incorporate uncertainty which is inherent to the ontology mapping process, and utilize the Dempster-Shafer model for dealing with incomplete and uncertain information produced during the mapping and combination in order to improve the correctness of the mapping. Our main objective was to assess how applying the belief function can improve correctness of the ontology mapping through combining the similarities which were originally created by both syntactic and semantic similarity algorithms. We have participated and carried out experiments with the data sets of the Ontology Alignment Evaluation Initiative 2006 which served as a test bed to assess both the strong and weak points of our system. The experiments confirm that our algorithm performs well with both concept and property names.
The combination of different similarity methods in the current ontology mapping approaches can considerably increase the quality of the mappings however uncertainty caused by incomplete or inconsistent data has received relatively little attention in the ontology mapping community. This paper describes a framework for integrating similarity measures and Dempster-Shafer belief functions for ontology mapping in the context of multi agent ontology mapping. Our novel approach describes how to incorporate uncertainty which is inherent to the ontology mapping process, and utilize the Dempster-Shafer model for dealing with incomplete and uncertain information produced during the mapping and combination in order to improve the correctness of the mapping. Our main objective was to assess how applying the belief function can improve correctness of the ontology mapping through combining the similarities which were originally created by both syntactic and semantic similarity algorithms. We have participated and carried out experiments with the data sets of the Ontology Alignment Evaluation Initiative 2006 which served as a test bed to assess both the strong and weak points of our system. The experiments confirm that our algorithm performs well with both concept and property names.
Click below to play the event (43 minutes)