Mining of negative association rules using improved frequent pattern tree

Extraction of interesting negative association rules from large data sets is measured as a key feature of data mining. Many researchers discovered numerous algorithms and methods to find out negative and positive association rules. From the existing approaches, the frequent pattern growth (FP-Growth) approach is well-organized and capable method for finding the item sets which are frequent, without the generation of candidate item sets. The drawback of FP-Growth is it discovers a huge amount of conditional FP-Tree. We propose a novel, improved FP-Tree for extracting negative association rules without generating conditional FP-Tree.