Implementasi Data Mining Dalam Menentukan Penjualan Paket Internet Dan TV Kabel Dengan Metode Algoritma C4.5 Pada PT Link Net

Abstract

APT Link Net is a multimedia company that became the first provider in Indonesia that provides internet speeds of up to 100 Mbps and provides more than 90 local and international channels with digital picture quality and stereo sound, PT. Link Net offers several choices of internet and cable TV packages such as Dlite, Elite, Family, Family Plus, Maxima, School, Supreme, judging by the large number of sales for these packages and there is no predictor of determining the sales of the best-selling and not-selling packages. The research uses data mining techniques using the C4.5 algorithm to analyze or determine package sales so that they can prepare more supporting tools to be used for best-selling packages. This study uses sales data sourced directly from PT. Link Net in 2021. The stages in the C4.5 Algorithm use six KDD (Knowledge Discovery in Databases) steps which include activities, namely, selection, preprocessing, transformation, data mining, Interpretation Evaluation. Through testing using rapidminer with an accuracy rate of 99.5%, thus the C4.5 Algorithm is very well used to implement this sales data. From the explanation above, it can be concluded that the best-selling Internet and cable TV packages at PT. Link Net namely Family and Family Plus.

Keywords: Data Mining, C4.5 Algorithm, Decision Tree, Prediction