Analyzing Customer Sentiment in Drug Reviews Using Natural Language Processing

Abstract

The continuous expansion of the internet has led to a surge in user-generated content. Patients frequently share feedback about medications they have taken to express their experiences and raise public awareness. The application of sentiment analysis plays a crucial role in the healthcare sector, as it allows for examining these public reviews to assess the efficacy and popularity of various medications. This research presents a comprehensive sentiment analysis of customer reviews from DrugRev, a dataset of drug reviews. The study evaluates customer sentiment regarding drug effectiveness, side effects, and ease of use. We employ natural language processing using multiple machine learning and deep learning algorithms to obtain valuable insights into customer satisfaction and preferences, aiding consumers and healthcare professionals in making informed medication choices. The findings demonstrate that Bidirectional Gated Recurrent Unit and BERT outperformed all other models with an F1 Score of 92.17. The outcomes of our experiments illustrate that our system can deliver precise, effective, and scalable drug suggestions. By understanding customer sentiments, we contribute to improving healthcare decision-making and enhancing the quality of healthcare services.