Open access Jurnal SINTA 2
Segmentasi Pelanggan Majalah pada Situs Web E-Commerce dengan K-Means++ dan Metode RFM
Jurnal Teknologi Informasi dan Ilmu Komputer (JTIIK) · Vol. 11 No. 6 (2024) · pp. 1243–1252
- ISSN
- 2528-6579, 2355-7699
- Published
- 2024-12-10
- Publisher
- Fakultas Ilmu Komputer, Universitas Brawijaya
- Language
- id
Abstract
Bahasa Indonesia
Segmentasi pelanggan merupakan salah satu metode yang dapat diterapkan untuk memaksimalkan peluang bisnis. Hal tersebut dapat membantu bisnis agar tetap kompetitif dalam persaingan pasar. Penerapan Artificial Intelligence (AI) dapat membantu dalam memberikan pemahaman kepada pelaku bisnis tentang segmentasi pelanggan berdasarkan riwayat transaksi. Penelitian ini menerapkan metode Recency, Frequency, and Monetary (RFM) yang dipadukan dengan algoritma clustering K-Means++ untuk melakukan segmentasi pelanggan. Silhouette score menjadi indikator pemilihan nilai k yang paling optimal dalam menentukan jumlah cluster. Kerangka kerja CRISP-DM yang digunakan dalam makalah ini juga membantu mempertahankan proses analisis yang konsisten. Pendekatan statistik sederhana ddigunakan untuk mengklasifikasikan setiap fitur dalam RFM menjadi label low, medium, dan high dalam hal menangkap pola segmentasi pelanggan. Hasil eksperimen menunjukkan nilai k = 3 sebagai yang paling optimal berdasarkan nilai WSS sebesar 843,214747 dan silhouette score sebesar 0,638181. Eksperimen juga menunjukkan bahwa cluster 0 memiliki nilai RFM rata-rata sebesar 1,14 (low), 1,20 (low), dan 301.640 (low). Cluster 1 memiliki nilai RFM rata-rata sebesar 249,61 (high), 2,62 (medium), dan 799,934 (medium). Cluster 2 memiliki nilai RFM rata-rata sebesar 233,01 (medium), 6,41 (high), dan 2018,088 (high). Abstract Customer segmentation is one method that can be applied to maximize business opportunities. It can help businesses remain competitive in the market competition. The application of Artificial Intelligence (AI) can assist in providing business stakeholders with an understanding of customer segmentation based on transaction history. This study applies the Recency, Frequency, and Monetary (RFM) method combined with the K-Means++ clustering algorithm for customer segmentation. The Silhouette score serves as an indicator for selecting the most optimal value of k to determine the number of clusters. The CRISP-DM framework used in this paper also helps maintain a consistent analysis process. A simple statistical approach is used to classify each RFM feature into low, medium, and high labels to capture customer segmentation patterns. Experimental results show that k = 3 is the most optimal value based on a WSS value of 843.214747 and a silhouette score of 0.638181. The experiments also indicate that Cluster 0 has average RFM values of 1.14 (low), 1.20 (low), and 301,640 (low). Cluster 1 has average RFM values of 249.61 (high), 2.62 (medium), and 799,934 (medium). Cluster 2 has average RFM values of 233.01 (medium), 6.41 (high), and 2018.088 (high).
English
Segmentasi pelanggan merupakan salah satu metode yang dapat diterapkan untuk memaksimalkan peluang bisnis. Hal tersebut dapat membantu bisnis agar tetap kompetitif dalam persaingan pasar. Penerapan Artificial Intelligence (AI) dapat membantu dalam memberikan pemahaman kepada pelaku bisnis tentang segmentasi pelanggan berdasarkan riwayat transaksi. Penelitian ini menerapkan metode Recency, Frequency, and Monetary (RFM) yang dipadukan dengan algoritma clustering K-Means++ untuk melakukan segmentasi pelanggan. Silhouette score menjadi indikator pemilihan nilai k yang paling optimal dalam menentukan jumlah cluster. Kerangka kerja CRISP-DM yang digunakan dalam makalah ini juga membantu mempertahankan proses analisis yang konsisten. Pendekatan statistik sederhana ddigunakan untuk mengklasifikasikan setiap fitur dalam RFM menjadi label low, medium, dan high dalam hal menangkap pola segmentasi pelanggan. Hasil eksperimen menunjukkan nilai k = 3 sebagai yang paling optimal berdasarkan nilai WSS sebesar 843,214747 dan silhouette score sebesar 0,638181. Eksperimen juga menunjukkan bahwa cluster 0 memiliki nilai RFM rata-rata sebesar 1,14 (low), 1,20 (low), dan 301.640 (low). Cluster 1 memiliki nilai RFM rata-rata sebesar 249,61 (high), 2,62 (medium), dan 799,934 (medium). Cluster 2 memiliki nilai RFM rata-rata sebesar 233,01 (medium), 6,41 (high), dan 2018,088 (high). Abstract Customer segmentation is one method that can be applied to maximize business opportunities. It can help businesses remain competitive in the market competition. The application of Artificial Intelligence (AI) can assist in providing business stakeholders with an understanding of customer segmentation based on transaction history. This study applies the Recency, Frequency, and Monetary (RFM) method combined with the K-Means++ clustering algorithm for customer segmentation. The Silhouette score serves as an indicator for selecting the most optimal value of k to determine the number of clusters. The CRISP-DM framework used in this paper also helps maintain a consistent analysis process. A simple statistical approach is used to classify each RFM feature into low, medium, and high labels to capture customer segmentation patterns. Experimental results show that k = 3 is the most optimal value based on a WSS value of 843.214747 and a silhouette score of 0.638181. The experiments also indicate that Cluster 0 has average RFM values of 1.14 (low), 1.20 (low), and 301,640 (low). Cluster 1 has average RFM values of 249.61 (high), 2.62 (medium), and 799,934 (medium). Cluster 2 has average RFM values of 233.01 (medium), 6.41 (high), and 2018.088 (high).
Keywords
Collaboration network of the authors
30 authors · 43 co-authorship links · 4 across different organisations
Hover or tap a name to highlight its co-authors. Double-click to open the author's page.
- Institut Teknologi Sepuluh Nopember
- University of Brawijaya
- Universitas Gadjah Mada
- Affiliation unknown
- Different organisations
- Same organisation
- Affiliation unknown
- Thicker line = more joint articles · bigger circle = more articles
Show as table
| Co-author | Organisation | Joint articles |
|---|---|---|
| Chastine Fatichah | Institut Teknologi Sepuluh Nopember | 3 |
| Achmad Arwan | University of Brawijaya | 2 |
| Suhadi Lili | Institut Teknologi Sepuluh Nopember | 2 |
| Dyah Sulistyowati Rahayu | Institut Teknologi Sepuluh Nopember | 1 |
| Bayu Priyambadha | University of Brawijaya | 1 |
| Aldy Sefan Rezanaldy | Affiliation unknown | 1 |
| Abdul Munif | Institut Teknologi Sepuluh Nopember | 1 |
| Vico Ade Candra | Affiliation unknown | 1 |
| Imam Kuswardayan | Institut Teknologi Sepuluh Nopember | 1 |
| Rahmi Ika Noviana | Affiliation unknown | 1 |
| Faizal Johan | Affiliation unknown | 1 |
| Widyanti Kartika | Affiliation unknown | 1 |
| Darlis Heru Murti | Institut Teknologi Sepuluh Nopember | 1 |
| Choiru Zain | Affiliation unknown | 1 |
| Kusbandono Ari Bowo | Affiliation unknown | 1 |
| Lutfi Rizal Gozali | Affiliation unknown | 1 |
| Ratih Nur Esti Anggraini | Institut Teknologi Sepuluh Nopember | 1 |
| Mohammad Ahmaluddin Zinni | Institut Teknologi Sepuluh Nopember | 1 |
| Fadli Husein Wattiheluw | Institut Teknologi Sepuluh Nopember | 1 |
| Lukman Hakim | Institut Teknologi Sepuluh Nopember | 1 |
| Ulima Inas Shabrina | Institut Teknologi Sepuluh Nopember | 1 |
| Muhammad Iskandar Java | Institut Teknologi Sepuluh Nopember | 1 |
| Divi Galih Prasetyo Putri | Universitas Gadjah Mada | 1 |
| Muhammad Shulhan Khairy | Institut Teknologi Sepuluh Nopember | 1 |
| Endang Wahyu Pamungkas | Institut Teknologi Sepuluh Nopember | 1 |
| Fernandes Sinaga | Institut Teknologi Sepuluh Nopember | 1 |
| Lukman Hakim | Institut Teknologi Sepuluh Nopember | 1 |
Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.