Implementasi Algoritma FP-Growth Untuk Menentukan Pola Pembelian Konsumen Pada Data Transaksi Penjualan PT Ligno Specialty Chemicals

Sari

Business competition in Indonesia is currently getting strict, especially in the
manufacturing industry. PT ligno Specialty chemicals is a chemical industry company that
applies the B2B model in their target market. PT Ligno must maintain relationships with
customers such as fulfilling customer demand for the stock of goods to be purchased.
However, determining the amount of inventory for sales stock is not an easy thing. At PT
Ligno in determining the amount of inventory is only based on mere estimates and on the
other side there is sales transaction data that has not been optimally utilized. So that an
alternative solution in this study is to use the data mining method with the FP-Growth
algorithm to find out the association rule between goods. This study aims to determine
consumer purchase patterns based on sales transaction data for 2018-2021 and evaluate the
results of association rules using lift ratios. The results showed that the FP-Growth algorithm
can be applied to determine consumer purchasing patterns in PT Ligno's sales transactions.
There are 8 association rules that meet the minimum support of 5% and the minimum
confidence of 80%, and consumers tend to buy items FGC-C043, FGC-C029, and FGC-C042
simultaneously, and also on items FGC-C012 and FGC-C013. The rules with the highest
support, confidence, and lift ratio values are if you buy FGC-C043 items, you also buy FGCC029
and FGC-C042 items with a support value of 6.3%, confidence of 100%, and lift ratio of
14,458. The results of the evaluation of the combination of minimum support and minimum
confidence, show that there are four rules with an lift ratio of less than . So that overall, the
rules generated in the FP-Growth algorithm in addition to the four rules are valid if they are
used as a reference in determining inventory.