Research on Improved Data-Mining Algorithm Based on Strong Correlation

The extensive application of association rules in commerce enables itself to be one of the most active research directions in data mining. Recently, the mining of strong correlation item pairs with statistical significance in transaction database receives a certain value. In order to further reduce the cost of testing candidate item pairs in relational database, we have developed the Taper algorithm according to 1NF property. The developed TaperR algorithm can cut the number of candidate pairs to improve effciency.