B0838
Title: Twitter sentiment analysis: Exploring users perceptions on health and well-being in Europe
Authors: Aurea Grane Chavez - Universidad Carlos III de Madrid (Spain)
Marco Zanotti - University of Milano-Bicocca (Italy)
Giancarlo Manzi - University of Milan (Italy) [presenting]
Qi Guo - Universidad Carlos III de Madrid (Spain)
Abstract: The goal is to identify the users' content on the Twitter social network for information related to health and well-being. The pandemic has changed people's thoughts, and people pay more attention to personal health and the national medical system. To track the evolution, an API tool is used in Python to scrap data from Twitter based on keywords such as \#long-term care, \#pension, \#insurance, and \#expectations for the future during a given period (before and after March 2019). Then, sentiment analysis is applied to selected tweets by different dimensions, including timeline, countries, languages, and related to local restriction policies. Retweets are also focused on, and statistical learning models are used to detect the potential pattern of keywords in tweets. Keywords can also identify users' attitudes. A global indicator, WBDI, was introduced to evaluate the health and well-being status of EU residents over 50. Regarding the WBDI indicator, health and well-being status levels are mapped by country. It presents that Northern EU countries have the best general status in 2018 and 2019; the distribution of WBDI is compared as a baseline to the performance of tweets.
