Sentiment Analysis of Traffic Congestion in Palembang Using Random Forest

Abstract

Traffic congestion is a persistent problem that significantly affects the daily activities of citizens in Palembang City. The public can voice their thoughts and worries about traffic conditions on social media, especially Facebook. This study uses the Random Forest algorithm to examine public opinion regarding traffic congestion in Palembang. The 2,021 Facebook comments in the dataset were gathered by web scraping and subjected to a number of preprocessing steps, such as cleaning, case folding, stemming, tokenization, normalization, and stopword removal. The TF-IDF algorithm was used for term weighting. The Random Forest model was trained and tested to classify sentiments into three categories: positive, neutral, and negative. The model attained good accuracy across training, testing, and validation datasets, according to the evaluation results. This research provides insights into public perceptions of traffic congestion and can serve as a reference for policymakers in developing data-driven strategies to address traffic issues in Palembang City.