Aspect-Based Sentiment Analysis on Social Media X for Electric Vehicles (EV) in Indonesia Using IndoBERT and Machine Learning

A countries around the world rallies in pushing more sustainable and more-environmentally-friendly approach in transportation vehicle development, the Ministry of Industry of Indonesia is stepping forward to be one the leader in transition towards environmentally friendly industries through the adoption of Electric Vehicles (EVs), EVs have become a prominent topic of discussion in Indonesia, particularly on social media platform X.Therefore, this study utilizes data from social media X to leverage Natural Language Processing (NLP) technology, focusing on sentiment analysis based on various aspects of EVs using the IndoBERT method. Previous research has found IndoBERT to be more effective compared to other Machine Learning methods. The use of NLP and IndoBERT is expected to provide a deep understanding of Indonesian public sentiment towards EVs, with analysis results compared across several Machine Learning and Deep Learning techniques. Model performance is evaluated using a confusion matrix, which provides metrics such as Accuracy, Precision, Recall, and F1 Score. The implementation of sentiment analysis based on multiple EV aspects is deployed through a website dashboard, facilitating easier access for users to gain insights and visualize sentiment-related data on EVs in Indonesia. Thus, this research not only contributes to enhancing public and industrial understanding of EVs but also opens up further potential for the development of Machine Learning and Deep Learning technologies for aspect-based sentiment analysis. From the research findings, the use of NLP and IndoBERT demonstrated superior performance compared to other methods, achieving a sentiment accuracy of 0,82 and aspect accuracy of 0,85. Negative sentiments predominated over positive ones, with negative sentiments dominating throughout the years 2021-2024, particularly concerning infrastructure aspects, and in 2024 regarding cost aspects.