Implementation of ECLAT Algorithm Technology: Determining Books Borrowing Pattern in University library

Abstract

Library is one place to get lots of information. Every transaction information recorded in a fairly large database, the large data if it can’t be used it will make the problem for the librarian. This research aims to determine the books that are often borrowed when borrowing, using the rules of the mining association using the ECLAT algorithm. Equivalence Class Transformation Algorithm (ECLAT) performs frequent itemset search from the bottom. This algorithm will only scan the data once, the scanning process will not be repeated to get frequent k-itemset. The database scan process in the ECLAT algorithm is not repeated because on the seacrh itemset not pay attention the sequence from the item. The result of this study indicate the performance of the ECLAT algorithm is good by requiring 15 ms in the execution process. 102 transaction data the same minimum support is 1%, resulting in 21 frequent itemset.