Application of Formal Concept Analysis in Association Rule Mining
Yong Liu, Xueqing Li · 2017
Data mining can find some interest information from large amounts of data. Data association (association rules) can find associations among data items. Data classification distinguishes every data from a data set or group, and it also can combine data association. Formal concept analysis is a data analyzing theory which discovers concept structure in data sets. It can transform formal context into concept lattice. This study applies association rules for classification based on formal concept analysis to classify the data. The proposed method creates concept lattice by using formal concept analysis, and generates association rules for classification from concept lattice. The rules will be pruned and sorted, and it will be used by following priority order. In order to estimate the performance of data classification, experiments have been done through a data set from UCI website. The evaluation indicators are correct rate and execute time. The result of experiments shows that the correct rate can increase after adjusting minimum support and minimum confidence.