Exploiting Social Media for Natural Language Processing: Bridging the Gap between Language-centric and Real-world Applications

Simone Paolo Ponzetto, Andrea Zielinski

Abstract: Social media like Twitter and micro-blogs provide a goldmine of text, shallow markup annotations and network structure. These information sources can all be exploited together in order to automatically acquire vast amounts of up-to-date, widecoverage structured knowledge. This knowledge, in turn, can be used to measure the pulse of a variety of social phenomena like political events, activism and stock prices, as well as to detect emerging events such as natural disasters (earthquakes, tsunami, etc.).