Analisis Sentimen Terhadap Review Film Menggunakan Metode Modified Balanced Random Forest dan Mutual Information

Firdausi Nuzula Zamzami (Universitas Telkom, Bandung)
Adiwijaya Adiwijaya (Universitas Telkom, Bandung)
Mahendra Dwifebri P (Universitas Telkom, Bandung)

JURNAL MEDIA INFORMATIKA BUDIDARMA
Vol 5, No 2 (2021): April 2021

Information exchange is currently the most happening on the internet. Information exchange can be done in many ways, such as expressing expressions on social media. One of them is reviewing a film. When someone reviews a film he will use his emotions to express their feelings, it can be positive or negative. The fast growth of the internet has made information more diverse, plentiful and unstructured. Sentiment analysis can handle this, because sentiment analysis is a classification process to understand opinions, interactions, and emotions of a document or text that is carried out automatically by a computer system. One suitable machine learning method is the Modified Balanced Random Forest. To deal with the various data, the feature selection used is Mutual Information. With these two methods, the system is able to produce an accuracy value of 79% and F1-scores value of 75%.