Socia media opinions aware adverse drug effect prediction and prevention system for the secured health care medical environment

Predicting adverse drug effects is difficult which is focused in various research works that focus on predicting the adverse effects of drugs based user reviews gathered from the social media. However those research works cannot accurately predict the user opinions side effects. These problems are resolved in the proposed research work by introducing the framework called “adverse drug effect aware drug recommendation system”. This research work focus on the online reviews about the drug reaction which is gathered from the twitter social media website. These reviews are analyzed to find the reaction of users in terms of positive reaction or negative reaction based on adverse effects. In this work training is done on the tweet data corpus downloaded from online to learn the negative and positive impact words. Initially preprocessing is done on the retrieved data reviews to eliminate unwanted words and result with only required data contents without noises and repeated data by using successor variety stemmer’s algorithm. After preprocessing, optimal feature selection is done on the preprocessed terms to select the most optimal terms that represent the drug reactions by using hybrid generic particle swarm optimization algorithm. Finally classification is done by using improved transductive support vector machine algorithm. The entire proposed work is simulated and analyzed in the matlab simulation environment from which it can be proved that the proposed research work tends to increased performance than the existing research methodologies.