ANALISIS TINGKAT KEPUASAN PENGGUNA APLIKASI SHOPEE MENGGUNAKAN ALGORITMA DECISION TREE

Abstract

The use of e-commerce applications like Shopee in Indonesia has increased significantly due to advancements in digital technology. Founded by Forrest Li in 2015, Shopee has become one of the leading e-commerce platforms in Southeast Asia, offering attractive features such as product discounts, cashback vouchers, and free shipping. This study aims to identify the factors influencing user satisfaction on Shopee to enhance competitiveness in the market. Data from a questionnaire survey includes service aspects such as efficiency, reliability, privacy, and responsiveness. Using the Decision Tree algorithm, the data is divided into training and testing sets, employing Cross Validation and Split Validation. The 10-Fold Validation shows the best performance with an accuracy of 93.78% and an AUC of 0.428. After optimization with feature selection and Split Validation, accuracy increases to 94.74% with an AUC of 0.653. These findings provide valuable guidance for Shopee developers in improving service quality and contribute significantly to research in the e-commerce field.

Penggunaan aplikasi e-commerce seperti Shopee di Indonesia meningkat signifikan berkat kemajuan teknologi digital. Didirikan oleh Forrest Li pada tahun 2015, Shopee kini menjadi salah satu platform e-commerce terkemuka di Asia Tenggara, menawarkan fitur menarik seperti diskon produk, voucher cashback, dan gratis ongkir. Penelitian ini bertujuan mengidentifikasi faktor-faktor yang memengaruhi kepuasan pengguna Shopee untuk meningkatkan daya saing di pasar. Data dari survei kuesioner mencakup aspek layanan seperti efisiensi, reliabilitas, privasi, dan responsivitas. Menggunakan algoritma Decision Tree, data dibagi menjadi training dan testing, serta menggunakan Cross Validation dan Split Validation. Validasi 10-Fold menunjukkan performa terbaik dengan akurasi 93.78% dan AUC 0.428. Setelah optimasi dengan fitur selection dan Split Validation, akurasi meningkat menjadi 94.74% dengan AUC 0.653. Temuan ini memberikan panduan berharga bagi pengembang Shopee dalam meningkatkan kualitas layanan serta kontribusi penting bagi penelitian di bidang e-commerce.