Providing an Effective Method to Increase the Accuracy of the Apriori Algorithm in Association Rules Mining

In general, the basis of data mining is the discovery of associative rules, the inductive algorithm being one of these algorithms and applied in all fields of data mining. The Enhanced BitApriori algorithm has somewhat improved the BitApriori algorithm by replacing some of the more effective techniques on the binary string. But this algorithm can still be improved by removing non-repetitive items. Removing unexpected items leads to a reduction in the number of candidates produced and, consequently, reducesmemory consumption, and therefore, the extracted rules are closer to the user s request and, if possible, lead to better management of the business. To be Therefore, it can be concluded that this could be of great importance in the investigation of this issue, and if this is to be done, it will be possible to innovate more effectively. An Enhanced BitAproiri algorithm for preseeding a set of item items first pruned a single member item and used the Trie structure to create a dual-member set of items. Therefore, for large bulk databases, the size of the Trie tree will increase, and so the memory will be used a lot. For this reason, in this study, we intend to improve the efficiency of this algorithm by providing a method. In order to verify the validity of the proposed method, the proposed algorithm is first compared with the Apriori classical algorithm, and then compared to the Enhanced BitApriori algorithm to evaluate the efficiency, run-time, memory consumption and number of candidate items of the proposed method. Finally, the results are evaluated by applying a dense and high volume data set. The results showed that the proposed method is more efficient.

کلیدواژه ها:

Effective Method ، Apriori Algorithm ، Increase the Accuracy ، Association Rules Mining