Predicting Drug Side-effects from Open Source Health Forums using Supervised Classifier Approach

Abstract

Adverse Drug Side-effects is an unexpected cause of drug reaction from patient to patient. However, Pharmaceutical companies manufacturing the drugs which are prescribed for the prevention and cure of the disease, sometimes leads to the ill effects and side-effects that leads to the death of consuming patients. Most of the side-effects are reported by pharmaceutical industries based on the Clinical trials out of which only a few of them are identified. The main aim of this paper is to analyze and predict the side-effects of the drugs which are massively available on public online health care forum by extracting various essential information such as side-effects, symptoms and medication of the consuming patients using UiPath tool from WebMB.com. Also, this research work uses Robotic Process Automation (RPA) to extract the data and pre-process manually by dropping noisy, unwanted records and missing data. Feature sets are analyzed for better predicting the Side-effects of the drugs. Machine Learning (ML) methods are used for performing a comparative analysis of Side-effects and identify which classifier gives better accuracy.