Text Classification on Sentiment Analysis of Marketplace SHOPEE Reviews On Twitter Using K-Nearest Neighbor (KNN) Method

Zulkifli Rais (Department of Statistics, Universitas Negeri Makassar, 90223, Makassar, Indonesia)
Rezky Novita Said (Department of Statistics, Universitas Negeri Makassar, 90223, Makassar, Indonesia)
Ruliana Ruliana (Department of Statistics, Universitas Negeri Makassar, 90223, Makassar, Indonesia)

JINAV: Journal of Information and Visualization
Vol. 3 No. 1 (2022)

This research aims to know the description and result of the classification sentiment analysis by Twitter users about Shopee. The method used in this research is K-Nearest Neighbor. K-Nearest Neighbor is a method that identifies groups or classifications based on the closest k of test data (training data). The most relative distance is calculated using the Euclidean distance. The data in this research were obtained from the Twitter API which used data on July 13, 2021, and the data according to the study were 150 tweets. Based on the results of the preprocessing text, there are 10 words that appear most often conveyed by Twitter users, and these opinions are related to the features provided by Shopee. The results obtained from this research are the highest level of text classification accuracy is 90% in training data and testing data comparison 80%: 20%