Missing values in Social Media: an application on Twitter data (Presentation)

The main aim of this work is to study the behaviour of users in social media analysis. In particular, the work involves Twitter users about the percepetion of the Italian guaranteed minimum income (”Reddito di cittadinanza”) on the basis of different categories of users. The main distinction about users is made between verified Twitter users and not verified users. The first category is related to politician, institutional authorities and other official stakeholders. The second one is represented by citizens and other subjects not directly involved in the process of realization of this measure. A classification method based on tweets, retweets and quotes posted by users with hashtag #redditodicittadinanza will be able to discern between verified and not verified users. Moreover, an analysis of the KPI (Key Performance Indicators) will be conducted using their presence and absence through the use of the complementary values. This tool is very useful to give a meaning to an absence of behaviour distinguishing between no interest and a negative opinion. Data have been collected in Italy in April 2019 using all tweets containing #redditodicittadinanza using the official Twitter API. Data will be analysed using R and Python. The use of this classification method allowed also to propose a model to recognize the presence of a Bot, a simulated Twitter user that login using same channels of the humans. Bot are often used to spread the effect of messages and topic in order to influence positive or negative effects.