Perbandingan Algoritma Naïve Bayes dan Decision Tree Pada Sentimen Analisis

Abstract

The development of information technology in recent years is fast, pushing internet usage. People use social media to share their opinions or interact simply with each other. Sentiment analysis is a branch of Natural Language Processing (NLP) that can filter and categorize people's views on social media. This study uses Twitter data to compare two algorithms: Naïve Bayes Classifier and Decision Tree. This study divided the data into two scenarios where the first one with 800 data and the second one with 200 data. Data from each scenario was divided again into 70% training data and 30% test data from 1000 data. The result shows that Naïve Bayes Classifier has much higher accuracy with 85% compared to the Decision Tree with 78% in the second scenario.