! Chévere! Text-Based Twitter Patterns from Venezuelan Food Shortages

摘要

Social media data from countries having challenges to free speech is a reliable form of journalism. An analysis is conducted to examine the social media response to the Venezuelan food shortages. Over 37,000 filtered Spanish tweets from the city of Caracas, Venezuela were used to observe reactions within each of the city's five municipalities. The number of tweets from December 2014 to October 2016 (23 months) is compared to the top trending tweet from July 19, 2017. Machine learning techniques show that certain tweets may be linked to a municipality within a 10 km radius. Tweet volume over almost two years indicates the significance of the shortages among the Venezuelan people engaged in the event.