Human Mobility Prediction Based on Social Media with Complex Event Processing

Abstract

The combination of mobile and social media sensors is foreseen to become a crucial course of action so as to comprehensively capture and understand the movement of people in large spatial regions. In that sense, the present work describes a novel personal location predictor that makes use of these two types of sensors. Firstly, it extracts the mobility models of an area capturing aspects related to particular users along with crowd-based features on the basis of geotagged tweets. Unlike previous approaches, the proposed solution mines such models in an online manner so that no previous off-line training is required. Then, on the basis of such models, a predictor able to forecast the next activity and position of a user is developed. Finally, the described approach is tested by using Twitter datasets from two different cities.

Fuzzy Modelling for Human Dynamics Based on Online Social Networks

“…Regarding location prediction, the anticipation of the future movement of a target individual is based on the idea that human mobility exhibits a high regularity, and, thus, predictability [ 6 ]. In this frame, our work also includes some innovative features with respect to existing literature related to OSN-based location predictors [ 16 , 56 , 57 , 58 ]. In this frame, most works make use of the spatio-temporal features of the documents in order to perform the prediction [ 56 , 57 ].…”

“…In this frame, our work also includes some innovative features with respect to existing literature related to OSN-based location predictors [ 16 , 56 , 57 , 58 ]. In this frame, most works make use of the spatio-temporal features of the documents in order to perform the prediction [ 56 , 57 ]. For example, like in [ 57 ], our approach also uses the spatio-temporal features of the documents of a user to make a prediction.…”

“…In this frame, most works make use of the spatio-temporal features of the documents in order to perform the prediction [ 56 , 57 ]. For example, like in [ 57 ], our approach also uses the spatio-temporal features of the documents of a user to make a prediction. However, the present work also takes under consideration the textual features of the documents in order to provide a prediction.…”

“…However, the present work also takes under consideration the textual features of the documents in order to provide a prediction. In addition to that, whilst [ 57 ] proposes a real-time system orchestrated by means of event-based rules, our work combines two steps: (i) an offline one to generate the clusters and the underlying mobility patterns and (ii) an online step where the patterns are used to generate a prediction in real time. This way, we avoid the convergence period problem that the mechanism in [ 57 ] suffers from.…”

“…In addition to that, whilst [ 57 ] proposes a real-time system orchestrated by means of event-based rules, our work combines two steps: (i) an offline one to generate the clusters and the underlying mobility patterns and (ii) an online step where the patterns are used to generate a prediction in real time. This way, we avoid the convergence period problem that the mechanism in [ 57 ] suffers from. Another difference exists regarding the particular prediction algorithm, while the work in [ 57 ] makes use of a Fallback Markov Model, we rely on fuzzy rules.…”