Abstract

Analyzing Urban Events and Dynamics with Big Data of Taxi Trips— A Tim
Track: Research & Sciences
Authors: xi zhu

we present a new approach for detecting and mapping urban events from big data of taxi OD trips based on location-specific time series decomposition and outlier detection. Our approach can distinguish different temporal trends , define and detect events as anomalies deviated from the trends, and map the dynamics of events over space and time with high accuracy and resolution.