Algoritmo narfo para mineração de regras de associação generalizadas não redundantes baseada em uma ontologia difusa
Abstract
Traditional approaches for mining generalized association rules are based only on database contents, and focus on exact matches among items. However, in many applications, the use of some background knowledge, as ontologies, can enhance the discovery process and generate semantically richer rules. In this way, this paper proposes the NARFO algorithm, a new
algorithm for mining non-redundant and generalized association rules based on fuzzy ontologies. Fuzzy ontology is used as background knowledge, to support the discovery
process and the generation of rules. One contribution of this work is the generalization of nonfrequent itemsets that helps to extract meaningful knowledge. NARFO algorithm also contributes at post-processing stage with its generalization and redundancy treatment.