ANALISIS SENTIMEN ULASAN KONSUMEN PADA PRODUK PONSEL PINTAR MENGGUNAKAN METODE NAÏVE BAYES

Abstract

In this modern era, people's sentiments or opinions are increasingly being disseminated and expressed freely in various media. Sentiment can be a great potential for companies who want to hear public feedback about their brand. Online shopping or e-commerce is a transaction carried out through the media in the form of online buying and selling websites or social networks that provide goods or services to be exchanged. An example of a well-known online shopping application in Indonesia is Shopee. Shopee has features that users can use to provide reviews, such as in the form of product star ratings or being able to comment on a product. On the Samsung Galaxy A03 Black product, there are reviews and ratings from various users who have bought and obtained the smartphone product. The purpose of checking this review data is to verify the sentiment classified by the machine learning model. The data used in this study is the review data of Samsung Galaxy A03 Black brand smartphone products sold online at Shopee Indonesia and managed directly by the Offical Samsung Shop. The data used amounted to 3007 data. Each text data that has been analyzed enters the scoring stage through the calculation process. The resulting sentiment analysis classification gets an accuracy and loss graph using Google Collaboratory, where giving reviews with testing and training data is in accordance with the rating of the reviews. The highest accuracy value shows 0.87 or equivalent to 87%, while the lowest accuracy value shows 0.81 or equivalent to 81%.