Identifying urban crowds using geo-located Social media data: a Twitter experiment in New York City

Abstract

The massive growth of GPS equipped smartphones coupled with the increasing importance of Social Media has led to the emergence of new services over LBSNs (Location-based Social Networks) where both, opinions and location, are shared. This proactive attitude allow us to consider citizens as sensors in motion whose information supports our approach: monitoring multitudes or crowds all around the city. More specifically, our proposal is mining geotagged data from LBSNs in order to analyze crowds according to different parameters as size, duration, composition, motivation, cohesion and proximity. This analysis is gathered under a methodology for crowd detection in cities that combines social data mining, density-based clustering and outlier detection into a solution that can operate on-the-fly. This methodology enables foreseeing crowds in short term based on the prior analysis of time and previous behavior of individuals in the geographical area under study. Our approach was validated using Twitter, as public social network par excellence, to analyze geotagged data in New York City on a normal day (reference day) and on New Year’s Eve, as the study day, when public crowds are expected.