How Twitter reveals Cities within Cities

Abstract

Cities are expanding and becoming more and more dynamic in terms of movement of population and, as a consequence, they are becoming melting pots with people of different cultures, religions and languages. In this paper, the authors use the multilingual analysis of Twitter to discover the "hidden cities" concealed within the city of Milan. Using the social media Twitter as a data source helps to detect weaker signals that are not captured through traditional census data. In this study, neighbourhoods in Milan are identified as areas where people speak mainly the same language on Twitter and these results are then compared with census data, to underline any parallelisms or discrepancies between the two sources of data. An added value of the paper is that the results are implemented within an online city dashboard, called Urbanscope.