Sentiment Analysis on Drug Reviews Using Ensemble Learning Method

Sentiment analysis, a branch of natural language processing, has gained considerable attention recently. This analytical approach is now widely used across multiple sectors, such as healthcare, finance, and customer service. Sentiment analysis in the healthcare domain can contribute to advancements in this field and help doctors to gain a deeper understanding of patient sentiments, enabling them to make more informed prescriptions. In our research, we concentrated on sentiment analysis of drug reviews. We applied ensemble learning methods, including the stacking learning method, using three base models and one meta-model. We also employed three feature extraction methods: TF-IDF, Bag of Words (BOW), and Continuous Bag of Words (CBOW). Among these methods, the stacking learning model with TF-IDF achieved the highest accuracy, reaching 89.8%.