Improvement Of Eclat Algorithm Based On Support In Frequent Itemset Mining

Finding frequent itemsets is computationally the most expensive step in association rules mining, and most of the research attention has been focused on it. With the observation that support plays an important role in frequent item mining, in this paper, a conjecture on support count is proved and improvements of traditional Eclat algorithm are presented. The new Bi-Eclat algorithm sorted on support: Items sort in descending order according to the frequencies in transaction cache while itemsets use ascending order of support during support count. Compared with traditional Eclat algorithm, the results of experiments show that the Bi-Eclat algorithm gains better performance on several public databases given. Furthermore, the Bi-Eclat algorithm is applied in analyzing combination principles of prescriptions for Hepatitis B in Traditional Chinese Medicine, which shows its efficiency and effectiveness in practical usefulness.