- Category: Framework Design/ System
- Project date: 2018-2020
With the widespread use of sensor-enabled smartphone devices, huge volume of GPS traces (e.g., timestamped location information) are generated facilitating various location-aware services. However, exploring spatio-temporal mobility dynamics (e.g., the intents of individuals’ movements or trip-purposes) is a challenging task as conventional information retrieval techniques fail to detect the underlying interpretations of movement history. This work focuses on how mobility information interprets the connection and correlations among different entities such as people, locations or point-of-interests (POIs), and other spatio-temporal contexts. Specifically, we propose a mobility analytics framework, called Mobilytics, to discover such correlations among individuals’ movement history using a novel mobility knowledge graph (MKG) and a deep-learning architecture that automatically annotates the GPS log. Further, a novel transfer learning technique is proposed to analyze the movement dynamics in another geographically dispersed region with the help of the knowledge gained from a region of similar type (say, academic campus). Experimental results on real-life datasets of two academic campuses demonstrate the efficacy of our proposed Mobilytics framework. Further, the analysis on huge volume of simulated traces (10,000 users) illustrates the scalability and robustness of Mobilytics framework.