Finding patterns between religions and emotions: A quantitative analysis based on twitter data

Abstract

The emotions someone associates with his or her religion and how this person talks about his or her faith have always been considered a personal topic. In this paper, the question of whether specific religions and emotions are connected is discussed. Based on Twitter data, individual networks, or so-called “tribes”, are created for four religions: Buddhism, Christianity, Islam and Judaism and four emotions: anger, fear, joy and sadness. Similarities and differences between tribes are analyzed using the content of the tweets. A network analysis is done for all tribes and the resulting data is used to create a machine learning model for each category. Using these, general patterns between emotions and religions are outlined and discussed. An analysis with further data was conducted on our model.