An Improved Graph Based Method for Extracting Association Rules
Wael Ahmad AlZoubi · International Journal of Software Engineering & Applications · 2015
This paper proposes an improved approach to mine strong association rules from an association graph, called graph based association rule mining (GBAR) method, where the association for each frequent itemset is represented by a sub-graph, then all sub-graphs are merged to determine association rules with high confidence and eliminate weak rules, the proposed graph based technique is self-motivated since it builds the association graph in a successive manner.These rules achieve the scalability and reduce the time needed to extract them.GBAR has been compared with three of the main graph based rule mining algorithms; they are, FP-Growth Graph algorithm, generalized association pattern generation (RIOMining) and multilevel association pattern generation (GRG).All of these algorithms depend on the construction of association graph to generate the desired association rules.On the other hand, this chapter expresses the observation results from the implementation of GBAR method recorded through the experiment.The detailed results are shown by different case studies in different minimum support thresholds values ranging from 90% down to 10% and minimum confidence values range from 55% to 95%.Generally, the observations focused on the execution time, the dimensionality of rules and the number of rules generated, because the performance of the association rule mining process affected directly of these criteria.Generally, the GBAR method has successfully reduced the execution time required to generate desired association rules in almost all of the dataset.