Identifying Adverse Drug Reactions By Analyzing Twitter Messages

Abstract

An Adverse Drug Reaction(ADR) is an unintended, harmful reaction resulting from an involvement of a medicinal product. Currently it has become the most common cause of deaths in the world, despite of post marketing drug surveillance. Expensive clinical trials do not uncover all the ADRs and also cumbersome for consumers and healthcare professionals. The majority of existing methods rely on patients' spontaneous self-reports. But these reporting systems are voluntary in most countries. The recent explosion of micro blogging platforms such as Twitter presents a new information source to discover ADRs. It has an enormous number of users who are frequently sharing and discussing their personal experiences, opinions, thoughts, and random details of their lives. In this study, we developed a system to automatically extract ADRs from Twitter messages utilizing Natural Language Processing(NLP) techniques. First, we proposed a novel method to lter out all the drug related messages including misspelled drug names from the Twitter data stream. Dictionary based approaches were used to identify medical terminology, emoticons and slang words. The interpretation of \internet language" was also address in this research. We evaluated our system on a manually annotated set of Twitter messages. The best classi er for the classi cation of ADR reached an accuracy of 68% with Fmeasure of 69%. The results suggest that there is potential for extracting ADR related information from Twitter messages to support pharmacovigilance.