Authors:
Marek Jaszuk
;
Teresa Mroczek
and
Barbara Fryc
Affiliation:
University of Information Technology and Management, Poland
Keyword(s):
Semantic Class, Automatic Ontology Building, Clustering Methods.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Artificial Intelligence
;
Biomedical Engineering
;
Collaboration and e-Services
;
Computational Semantics
;
Data Engineering
;
e-Business
;
Enterprise Information Systems
;
Health Information Systems
;
Information Integration
;
Information Retrieval
;
Information Systems Analysis and Specification
;
Integration/Interoperability
;
Knowledge Engineering and Ontology Development
;
Knowledge Management
;
Knowledge-Based Systems
;
Ontologies and the Semantic Web
;
Pattern Recognition
;
Society, e-Business and e-Government
;
Software Engineering
;
Symbolic Systems
;
Web Information Systems and Technologies
Abstract:
The paper discusses the problem of discovering semantic classes which are the basic building block of any semantic model. A methods based on clustering techniques is proposed, which leads to discovering related data coming from survey questions and other sources of information. We explain how the questions can be interpreted as belonging to the same semantic class. Discovering semantic classes is assumed to be foundation for construction of the knowledge model (ontology) describing objects being the subjects of the survey. The ultimate goal of the research is developing a methodology for automatic building of semantic models from the data. In our case the surveys refer to different socio-economic factors describing student’s situation. Thus the particular goal of the work is construction of the knowledge model, which would allow for predicting the possible outcomes of the educational process. The research is, however, more general, and its results could be used for analyzing collecti
ons of objects, for which we have data coming from surveys, and possibly some additional sources of information.
(More)