Application Of Naive Bayes Algorithm For Sentiment Analysis On Economic Recession Threat

Abstract

Recession is a condition in which real economic growth becomes negative, or in other words, there is a decline in Gross Domestic Product (GDP) for two consecutive quarters in one year. A recession is characterized by a weakening of the global economy that has an impact on the domestic economy in various countries. The greater the dependence of a country on the global economy, the more likely the country is to experience a recession. An economic recession can cause a simultaneous decline in all economic activities, including corporate profits, employment, and investment. In this study, data was collected from YouTube using a crawling technique, with a total of 200 comments analyzed. These comments were then labeled with a lexicon-based method using an Indonesian dictionary. The preprocessing stage was carried out to prepare the data before sentiment analysis. In addition, the TF-IDF word weighting method was applied with the bigram feature (n = 1) in the analysis. The system was evaluated using a confusion matrix, and the results showed that the prediction model, which was based on 200 opinion data with a 9:1 split ratio between training data and test data, achieved an accuracy of 75.00%. However, the precision, recall, and F1-score values each show 0.00%. The performance of the system model built in this study shows less than satisfactory results and may require improvements to increase its effectiveness.