An application of people's sentiment from social media to smart cities

Abstract

This paper presents a method for analyzing social media and digital governance in smart cities. This research, after selecting a group of issues related to people's concerns about daily life in cities, makes use of Afinn lexicon as dictionary with sentiment score in order to recognize and measure the strength of individual words. The experiment retrieves different tweets classified by hashtags in a database and calculates an average sentiment. It involves some statistical processes for the recognition of the topics' sentiment: the calculation of the weighted average, the standard deviation, and the distribution among -5 to +5, based on the score range used by the lexicon. Finally, human evaluation is used to compare results from the automatic system.