The time- and geo-coordinates associated with a sequence of tweets manifest the spatial-temporal movements of people in real life. This paper aims to analyze such movements to predict the next location of an individual based on the observations of his mobility behavior over some period of time and the recent locations that he has visited. To this end, we defined a prediction methodology based on a set of spatio-temporal features characterizing locations and movements among them. We then combined the features in a supervised learning approach based on M5 model trees. The experimental results obtained by using a real-world dataset show that the supervised method is effective in predicting the users next places achieving a remarkable accuracy.

Published in: International Conference on Information Society (i-Society 2017)

• Date of Conference: 17-19 July 2017
• DOI: 10.2053/iSociety.2017.0024
• ISBN: 978-1-908320-80-3
• Conference Location: Dublin, Ireland