Analysis and prediction of influenza in the UAE based on Arabic tweets

Social Networks have been used recently for detecting different events, such as diseases. In this paper, we propose a system to analyze and predict influenza based on Arabic Twitter data. The proposed systems filters and extracts features from the Arabic tweets and classifies them into three categories; self-reporting, nonself-reporting and non-reporting. The classified tweets are then used to analyze the spread of Influenza in the UAE. Moreover, the classified tweets are used to predict the number of future hospitals visits by feeding the tweets into a linear regression model. The conducted evaluation show high correlation between the number of analyzed tweets and the number of hospital visits obtained from the Ministry of Health, UAE, which demonstrates the feasibility of the proposed system.