Document Type : Original Article
Faculty of Computer Engineering and IT, Shahrood University of Technology, Shahrood, Iran
Faculty of Computer and IT Engineering, Mazandaran University of Science and Technology, Babol, Iran
Sydney International School of Technology and Commerce, Sydney, Australia
Election prediction has always been of interest to many people. In the last decade, an increasing influence of social networks and the possibility of sharing opinions and ideas has rendered election prediction based on social network data analysis. This paper, drawing on Twitter data and sentiment analysis, uses the proportion of positive messages rate to negative messages rate as an effective indicator for predicting elections. Then, using the aging estimation method, it predicts the values of this indicator in future time windows. The experiments conducted on Twitter data related to the 2020 United States presidential election in a four-month time window indicate that the indicator values and eventually the election results can be predicted with high accuracy.