Spatio-temporal modeling of city events combining datasets in cyberspace and real space

The scale and impacts of city events, including cultural, entertainment and sporting events, reflect the economics and culture of a city to a certain extent. The occurrence of events affects urban city significantly both online in cyberspace and offline in real space. The evolutionary perception, dynamic modeling and spatio-temporal analysis of city events from cyberspace and real space, are of important theoretical research and application value. This article proposes a novel approach of city events spatio-temporal modeling and analysis with trajectories and social media datasets in real space and cyberspace respectively. The approach firstly identifies statistically significant anomalous city regions and traffic flows with trajectories during the events, analyzing spatio-temporal change of events in real space, then analyzes spatio-temporal change during the whole process of city events in cyberspace. Finally, this article presents a modeling approach for characterizing the development and evolution of urban geospatial and behavioral space throughout events, with datasets in cyberspace and real space combined. Taking the Opus II Jay World Tour in 2015 as an example, employing taxi GPS trace data and Weibo data in Wuhan, the proposed method realizes the whole process modeling and evolution analysis of urban geospatial and behavioral space before, during and after the event. Then the method is compared with two other approaches that use either real space dataset or cyberspace dataset alone. The experimental results show that the proposed approach measures the impact of an event both in cyberspace and real space with reason, and describes city events effectively. © 2019, Surveying and Mapping Press. All right reserved.