Title: Analysis of drug reviews using data analytics and sentiment analysis

Authors: Bijayalaxmi Panda; Chhabi Rani Panigrahi; Bibudhendu Pati

Addresses: Department of Computer Science and Technology, Gita Autonomous College, Bhubaneswar, India ' Department of Computer Science, Rama Devi Women's University, Bhubaneswar, India ' Department of Computer Science, Rama Devi Women's University, Bhubaneswar, India

Abstract: This study endeavours to manage and analyse the volume of data in our daily lives by employing a drug review dataset from the UCI machine learning library. Focusing on six key parameters - drug-id, name, condition, review, rating, and usefulness count - we meticulously curated a subset of dataset, targeting the top ten conditions. Our approach integrates sentiment analysis (SA) and exploratory data analysis (EDA) to glean insights. The EDA provides an overview, including details on the top ten drugs for each condition, their respective value counts, mean ratings, mean polarity, the number of reviews, and quantity of medications. Our experimentation reveals phentermine as the most popular and efficacious drug for obesity, based on mean rating and polarity, while ethinyl estradiol or norethindrone emerges as the least favoured drug by rating, and Skyla by mean polarity. This work aims to contribute valuable insights into drug efficacy and popularity, facilitating informed decision-making in healthcare.

Keywords: drug review; exploratory data analysis; EDA; sentiment analysis; polarity; rating; machine learning.

DOI: 10.1504/IJCSE.2025.149767

International Journal of Computational Science and Engineering, 2025 Vol.28 No.6, pp.619 - 627

Received: 26 Oct 2023
Accepted: 16 Feb 2024
Published online: 12 Nov 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article