By Sven Reulen (Netherlands)
The goal of this research is to analyze links between the sentiments of citizens and the (spatial and social effects of) demographics in shrinkage regions. These regions are experiencing the demographic developments population shrinkage, aging population, declining working population and a changing household composition. Citizens are the best observers of the social and spatial effects of a changing demographic. They could be a useful participating party in guiding the changes an area is going through. In order to measure the sentiment of citizens, twitter messages (tweets) are analyzed. Municipalities where shrinkage is indicated are compared to non-shrinking municipalities. Tweets are converted in such a way that the tweets can be explored and related to the level of shrinkage of the municipalities. In order to link tweets to the sentiment of a population three possible relations are set up. The first relation aims at the type of locations mostly tweeted about in a municipality. The second relation inspects tweets containing words that indicate spatial effects of shrinkage. As a third relation the overall sentiment is grasped by executing a questionnaire, and letting people categorize the gathered tweets in to sentiment categories. Relations between the content of tweets and the shrinkage level indicate that there are differences in sentiment in different municipalities. A part of these differences appear to be related to shrinkage. In municipalities that are characteristic to shrinkage regions (shrinkage municipalities), people more often tweet about smaller scaled locations. Also, there is less diversity in locations tweeted about. The second relation indicates that more tweets in shrinkage municipalities are about vacant buildings and forced movings. The third relation indicates that people in shrinkage municipalities post tweets that in general are more negative, more about the past and less about themselves. The relations found between sentiment and shrinkage give insight in the sentiment of the population in shrinkage regions compared to the sentiment in other municipalities. For governments, organizations and citizens understanding the sentiment can be of importance. Policies can be set up, supported or criticized that aim on these regions. Understanding the sentiment of a population could also influence the finding of a location to start or close an establishment, or to find a location to live. Additionally the methods executed to derive the sentiment of a population by exploring tweets can give insight in to the increasing amount of research that includes online communication.
Keywords: Shrinkage regions; social media; georeferenced tweets; sentiment; Netherlands; Social Media Analytics (SMA).