Frequent Pattern Growth for Predicting the Pattern of Office Stationery Needs

Paska Marto Hasugian (Software Engineering Study Program, STMIK Pelita Nusanatara)
Fenius Halawa (Software Engineering Study Program, STMIK Pelita Nusanatara)
Dharma Rajen Kartighaiyab (Software Engineering Study Program, STMIK Pelita Nusanatara)

Jurnal Mantik
Vol. 5 No. 4 (2022): February: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)

One of the supporters of operational activities from the agency is ATK (Office Stationery), work will be constrained if these components are not available, as well as STMIK Pelita Nusantara requires ATK to be used every day starting from study programs, institutions, finance and other units. The problem in this research is that the accumulation of ATK purchase data has never been used or analyzed. With this problem, the researcher wants to perform an extraction on the available data so that knowledge is found in the form of predictions in the form of patterns of ATK needs. For the problem solving process, Frequent Pattern-Growth (FP-Growth) is used. FP-Growth is an alternative algorithm that can be used to determine the data set that appears most frequently (frequent itemset) in a data set. The FP-Growth algorithm is an algorithm that is very efficient in searching for frequent itemset. FP-Growth uses a different approach from the algorithm that is often used, namely the a priori algorithm. This algorithm stores information about frequent itemset in the form of FP-Tree. The FP-Tree that is formed can take advantage of transaction data that has the same item so that it can reduce repeated database scans in the mining process and can take place more quickly. FP-Growth uses a different approach from the algorithm that is often used, namely the a priori algorithm. This algorithm stores information about frequent itemset in the form of FP-Tree. The FP-Tree that is formed can take advantage of transaction data that has the same item so that it can reduce repeated database scans in the mining process and can take place more quickly. FP-Growth uses a different approach from the algorithm that is often used, namely the a priori algorithm. This algorithm stores information about frequent itemset in the form of FP-Tree. The FP-Tree that is formed can take advantage of transaction data that has the same item so that it can reduce repeated database scans in the mining process and can take place more quickly..