E-COMMERCE WEBSITE APPLICATION WITH CUSTOMER LOYALTY AND RECOMMENDATIONS ITEMS DEPENDS ON PRICE FEATURES USING K- MEANS CLUSTERING METHOD

Abstract

In the current era of computer technology development, the internet is a media of information that can be accessed across countries. To overcome problems in the commercial system, E-Commerce was created. E-Commerce is a commercial activity using internet media. In the midst of the pandemic Covid-19, many Indonesians have opened E- Commerce to fullfill their daily needs because they have been affected by this Covid-19 pandemic. This program is one of the forms of E-Commerce that was created during the Covid-19 pandemic, starting from December 2020 which was named Maniak Mainan at Indonesian E-Commerce, Tokopedia. Currently the products listed on Tokopedia Maniak Mainan is around 6100 products, therefore a recommendation program is made to provide users with suitable choices according to the price chosen through the bottom of the product page. This program also has a customer loyalty feature that allows users to get discounted prices by adding the nominal amount of purchases on the Maniak Mainan website. The recommendations and customer loyalty made using the K-Means Clustering method obtained using original sales and product data from Tokopedia Maniak Mainan. However, this cluster center will adjusted when the admin uploads a product with a higher or lower price, and the customer loyalty tier will also change if the user's purchase nominal exceeds the previous highest nominal. Through the results of the test, it can be concluded that 20 cluster centers are the right distance for this because it includes the price of products that are nearby. The compatibility test was carried out with the Silhouette Score with the results of 0.579 on the recommendation and 0.689 on the customer tier