Please use this identifier to cite or link to this item:
https://repositori.uma.ac.id/handle/123456789/27332
Title: | Analisis Sentimen Produk Berdasarkan Review Pelanggan Shopee Menggunakan KNN |
Other Titles: | Sentiment Analysis of Shopee Product Reviews Using the K-Nearest Neighbors (KNN) Algorithm |
Authors: | Irwannia, Fira |
metadata.dc.contributor.advisor: | Lubis, Andre Hasudungan |
Keywords: | Analisis Sentimen;Ulasan Produk;KNN;Shopee;TF-IDF;Sentiment Analysis;Product Reviews |
Issue Date: | 2025 |
Publisher: | Universitas Medan Area |
Series/Report no.: | NPM;188160048 |
Abstract: | Penelitian ini bertujuan untuk melakukan analisis sentimen terhadap ulasan pelanggan mengenai produk mukena yang tersedia di aplikasi Shopee menggunakan algoritma K-Nearest Neighbors (KNN). Data yang digunakan merupakan data primer sebanyak 200 ulasan yang dikumpulkan secara manual. Proses analisis dimulai dari preprocessing data berupa case folding, tokenisasi, penghapusan stopwords, dan stemming, kemudian dilanjutkan dengan ekstraksi fitur menggunakan metode TF-IDF, dan klasifikasi menggunakan algoritma KNN. Penelitian ini juga melakukan evaluasi terhadap performa model dengan akurasi. Hasil pengujian menunjukkan bahwa proporsi data training dan nilai parameter n_neighbors sangat mempengaruhi akurasi model. Proporsi data training sebesar 90% dan testing 10% menghasilkan akurasi tertinggi sebesar 90%. Namun, ketika nilai n_neighbors = 3, proporsi 70:30 menghasilkan performa terbaik sebesar 81,67%. Penelitian ini menunjukkan bahwa KNN mampu digunakan sebagai metode yang efektif dalam analisis sentimen terhadap ulasan produk. This study aims to conduct sentiment analysis on customer reviews of mukena products available on the Shopee application using the K-Nearest Neighbors (KNN) algorithm. The data used is primary data consisting of 200 reviews collected manually. The analysis process begins with data preprocessing such as case folding, tokenization, stopword removal, and stemming, followed by feature extraction using the TF-IDF method, and classification using the KNN algorithm. The model's performance is evaluated using a confusion matrix. The results show that the proportion of training data and the n_neighbors parameter significantly affect the model's accuracy. A 90% training and 10% testing proportion produced the highest accuracy of 90%. However, with n_neighbors = 3, the best performance was achieved with a 70:30 data split, reaching 81.67% accuracy. This study demonstrates that KNN is an effective method for sentiment analysis on product reviews. |
Description: | 12 Halaman |
URI: | https://repositori.uma.ac.id/handle/123456789/27332 |
Appears in Collections: | SP - Informatic Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
188160048 - Fira Irwannia - Fulltext.pdf | Fulltext | 1.48 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.