| Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 92 === By storing the transactional database in an in-memory data structure, we can speed up the task of association rule mining. In this paper, we present a new algorithm for association rule mining. In our algorithm, we store the transactional database in a list called the frequent list. By manipulating the frequent list, the algorithm produces all the frequent itemsets. We implemented some existing representative algorithms for association rules. We then performed benchmarking on these algorithms using synthetic databases. The benchmarking shows that when the minimum support is small or when the size of the database is large, cases that require extensive computation, our algorithm outperforms the existing algorithms. When the minimum support is large or when the size of the database is small, the difference in execution time of our algorithm and other algorithms is not significant.
|
|---|
