Claim Missing Document
Check
Articles

Found 1 Documents
Search

Pengelompokan Ulasan Produk HP pada Marketplace Tokopedia menggunakan Metode Semi Supervised K-Means Rizky Ardiawan; Yuita Arum Sari; Bayu Rahayudi
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 6 No 1 (2022): Januari 2022
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The internet has grown rapidly in accordance with the changing times. It also changes shopping behavior that was originally face to face now can be done online. Cell phones or smartphones are the most sought after items today. To buy these items online, there are many marketplaces available in Indonesia, such as Tokopedia. A product review is rated as the main factor for consumers to buy goods. To perform analysis on reviews, a method is needed that can classify and group reviews into existing categories. By combining the two understandings between Supervised and unsupervised, one can create a grouping method based on training data consisting of labeled data. The method that is suitable for this case is the Semi Supervised K-Means method. From the results of this study, it was found that in 4 different experiments, the evaluation of the cluster value using Silhouette was 0.013647 which was the largest value using the Semi Supervised K-Means method. Which is very small, namely 3 clusters. However, the results of clustering the clusters produced in the same method proved to be better than the K-Means method in general with the review data according to the original label.