Analisis Sentimen Ketidakstabilan Harga Gabah Berbasis Data Twitter

Abstract

- Public opinion, whether positive, negative, or neutral, regarding a particular policy or phenomenon in society, is a valuable thing to analyze through a method known as sentiment analysis. The case in this study is the decline in grain prices in early to mid-2021. This study aims to determine the percentage of sentiment polarity that appears when associated with the keyword price of grain and determine the level of accuracy of sentiment class predictions using the Naïve Bayes method. The results showed that the largest percentage of sentiment was negative as much as 46.30%, neutral 32.70% and positive as much as 20.99%. The results of the wordcloud also show that twitter users link the issue of grain prices to rice imports, the role of the government and fertilizers. The results of the classification show a fairly good accuracy value of 67.32%.