Enhancement of association rules interpretability using generalization

Abstract EN

-Data mining has a number of common methods.

One of such methods is the association rules mining.

While association rules mining often produces huge number of rules, it prevents the analyst from finding interesting rules and consequently, this method is a waste of time.

Visualization is one of the methods to solve such problems.

However, most of the association rule visualization techniques are suffering from viewing huge number of rules.

This paper provides a modification on the techniques of the visualization to help the analyst to interpret the association rules by grouping the large number of rules using a modified Attribute Oriented Induction algorithm, then; these grouped rules are visualized using a grouped graph method.

Experimental results show that the proposed technique produces excellent compression ratio.