Daniel Theodorus
Unknown Affiliation

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Machine Learning Rekomendasi Produk dalam Penjualan Menggunakan Metode Item-Based Collaborative Filtering Daniel Theodorus; Sarjon Defit; Gunadi Widi Nurcahyo
Jurnal Informasi dan Teknologi 2021, Vol. 3, No. 4
Publisher : SEULANGA SYSTEM PUBLISHER

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37034/jidt.v3i4.151

Abstract

The shift towards Industry 4.0 has pushed many companies to adopt a digital system. With the sheer amount of data available today, companies start to face difficulties with providing product recommendation to their customers. As a result, data analysis has become increasingly important in the pursuit of providing the best service (user experience) to customers. The location appointed in this research is PT. Sentral Tukang Indonesia which is engaged in the sale of building materials and carpentry tools such as: paint, plywood, aluminum, ceramics, and hpl. Machine Learning has emerged as a possible solution in the field of data analysis. The recommendation system emerged as a solution in providing product recommendation based on interactions between customers in historical sales data. The purpose of this study is to assist companies in providing product recommendation to increase sales, to make it easier for customers to find the products they need, providing the best service (user experience) to customers. The data used is customer, item, and historical sales at PT. Sentral Tukang Indonesia over a time span of 1 period.data historical sales divide to dataset training 80% and dataset testing 20%. The Item-based Collaborative Filtering method used in this study uses Cosine Similarity algorithm to calculate the level of similarity between products. Score prediction uses Weighted Sum formula while computation of error rate uses the Root Mean Squared Error formula. The result of this study shows top 10 product recommendations per customer. The products displayed are products with the highest score from the individual customer. This research can be used as a reference by companies looking to provide product recommendations needed by their customers.