Authors:
Rafael Garcia Miani
1
;
Cristiane Akemi Yaguinuma
2
;
Marilde Terezinha Prado Santos
2
and
Vinícius Ramos Toledo Ferraz
2
Affiliations:
1
IBM Brazil Software Laboratory, Brazil
;
2
Federal University of São Carlos, Brazil
Keyword(s):
Data Mining, Generalized Semantic Association Rules, Redundant Rules, Fuzzy Ontology.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Biomedical Engineering
;
Business Analytics
;
Data Engineering
;
Data Mining
;
Databases and Information Systems Integration
;
Datamining
;
Enterprise Information Systems
;
Health Information Systems
;
Information Systems Analysis and Specification
;
Knowledge Engineering and Ontology Development
;
Knowledge Management
;
Knowledge-Based Systems
;
Ontologies and the Semantic Web
;
Ontology Engineering
;
Sensor Networks
;
Signal Processing
;
Society, e-Business and e-Government
;
Soft Computing
;
Symbolic Systems
;
Web Information Systems and Technologies
Abstract:
This paper proposes the NARFO* algorithm, an algorithm for mining non-redundant and generalized association rules based on fuzzy ontologies. The main contribution of this work is to optimize the process of obtaining non-redundant and generalized semantic association rules by introducing the minGen (Minimal Generalization) parameter in the latest version of NARFO algorithm. This parameter acts on generalize rules, especially the ones with low minimum support, preserving their semantic and eliminating redundancy, thus reducing considerably the amount of generated rules. Experiments showed that NARFO* produces semantic rules, without redundancy, obtaining 68,75% and 55,54% of reduction in comparison with XSSDM algorithm and NARFO algorithm, respectively.