Research on Time-Validity and Incremental Association Rules

Association rule mining is an important research field in data mining. Current association-rule methods mainly depend on the support-confidence framework. The strategy is not quite effective in consideration of the time-sensitive factor and correlation problem between antecedent and consequent of rules. To solve this problem, a novel association rules framework has been proposed: time-validity support and time-validity match. First, we define a new match as the substitution of confidence for coping with the correlation problem between antecedence and consequence of rules. Moreover, by embedding the time-entropy factor into the new support-match framework, the time-sensitive can be solved. Finally, an example is given to prove the feasibility and superiority of the new method. On the basis of this, we propose a new incremental algorithm and the idea of performance. Experimental results and comparisons with traditional incremental method demonstrate the effectiveness of the proposed framework.