Nova Nurul Putri, - (2023) PERBANDINGAN METODE K-NEAREST NEIGHBOR (K-NN) DAN DECISION TREE PADA ANALISIS SENTIMEN ULASAN PENGGUNA APLIKASI EXPEDISI JNE DI GOOGLE PLAY STORE. S1 thesis, Universitas Pendidikan Indonesia.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_SISTEL_1909291_Title.pdf Download (498kB) |
|
Text
S_SISTEL_1909291_Chapter1.pdf Download (280kB) |
|
Text
S_SISTEL_1909291_Chapter2.pdf Restricted to Staf Perpustakaan Download (1MB) |
|
Text
S_SISTEL_1909291_Chapter3.pdf Download (257kB) |
|
Text
S_SISTEL_1909291_Chapter4.pdf Restricted to Staf Perpustakaan Download (971kB) |
|
Text
S_SISTEL_1909291_Chapter5.pdf Download (168kB) |
|
Text
S_SISTEL_1909291_Appendix.pdf Restricted to Staf Perpustakaan Download (335kB) |
Abstract
Perkembangan teknologi internet telah mengubah lanskap bisnis dengan mendorong pertumbuhan e-commerce. E-commerce, sebagai platform penjualan produk dan layanan secara online, sangat tergantung pada industri logistik, khususnya layanan ekspedisi. Di Indonesia, layanan ekspedisi seperti JNE, J&T, dan Pos Indonesia memainkan peran penting dalam pengiriman produk dari penjual ke pelanggan. Namun, keberhasilan layanan ekspedisi juga tergantung pada kualitas dan kecepatan pengiriman, yang tercermin dalam ulasan pelanggan di platform seperti Google Play. Penelitian ini menganalisis ulasan pelanggan terkait layanan ekspedisi JNE menggunakan metode klasifikasi seperti K-Nearest Neighbors (KNN) dan Decision Tree. Data latih dan data uji dipisahkan untuk evaluasi model, dan performa diukur dengan akurasi, presisi, recall, dan F1-Score. Pada hasil evaluasi menggunakan KNN, ditemukan bahwa akurasi model berada pada kisaran 87,2% hingga 87,6% dengan metrik jarak Minkowski, Euclidean, dan Manhattan. Pada hasil evaluasi menggunakan Decision Tree, akurasi model berada pada kisaran 90,4% hingga 90,7% dengan kriteria Gini, Entropy, dan log_loss. Selain itu, visualisasi data dilakukan untuk mengetahui kata-kata yang paling sering muncul dalam ulasan positif dan negatif. Kata-kata seperti "lambat," "buruk," dan "pengiriman" sering muncul dalam ulasan negatif. Dari hasil perbandingan, ditemukan bahwa hasil akurasi pada model Decision Tree cenderung memberikan performa yang lebih baik dibanding KNN dalam mengklasifikasikan ulasan pengguna terkait layanan ekspedisi JNE. Secara keseluruhan, penelitian ini memberikan wawasan tentang bagaimana metode klasifikasi dapat digunakan untuk menganalisis ulasan pelanggan dan memahami sentimen terkait layanan ekspedisi. ----- The development of internet technology has changed the business landscape by driving the growth of e-commerce. E-commerce, as a platform for selling products and services online, is highly dependent on the logistics industry, especially shipping services. In Indonesia, forwarding services such as JNE, J&T and Pos Indonesia play an important role in delivering products from sellers to customers. However, the success of delivery services also depends on the quality and speed of delivery, which is reflected in customer reviews on platforms such as Google Play. This study analyzes customer reviews regarding JNE courier services using classification methods such as K-Nearest Neighbors (KNN) and Decision Tree. Training data and test data are separated for model evaluation, and performance is measured by accuracy, precision, recall, and F1-Score. In the evaluation results using KNN, it was found that the accuracy of the model was in the range of 87.2% to 87.6% with the Minkowski, Euclidean, and Manhattan distance metrics. In the evaluation results using the Decision Tree, the model accuracy is in the range of 90.4% to 90.7% with the Gini, Entropy, and log_loss criteria. In addition, data visualization was carried out to find out the words that appear most often in positive and negative reviews. Words like "slow," "poor," and "delivery" frequently appear in negative reviews, while words like "good," "service," and "convenient" appear in positive reviews. From the results of the comparison, it was found that the Decision Tree model tends to provide better performance than KNN in classifying user reviews regarding JNE expedition services. Overall, this study provides insight into how classification methods can be used to analyze customer reviews and understand sentiment regarding courier services.
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | hafiyyan putra pratama : https://scholar.google.com/citations?user=tQe1410AAAAJ&hl=id&oi=ao ahmad fauzi : https://scholar.google.com/citations?hl=id&user=b6BGJbEAAAAJ ID SINTA Dosen Pembimbing: hafiyyan: 6681148 Ahmad fauzi: 6122861 |
| Uncontrolled Keywords: | JNE, Analisis Sentimen, K-Nearest Neighbor (K-NN), Decision Tree |
| Subjects: | L Education > L Education (General) T Technology > T Technology (General) |
| Divisions: | UPI Kampus Purwakarta > S1 Sistem Telekomunikasi |
| Depositing User: | Nova Nurul Putri |
| Date Deposited: | 04 Sep 2023 02:57 |
| Last Modified: | 04 Sep 2023 02:57 |
| URI: | http://repository.upi.edu/id/eprint/101797 |
Actions (login required)
 |
View Item |