Green urban infrastructure, which serves the interests of both human and nature, are considered as essential assets of urban residents. However, measuring the use of green space has been problematic. Because, most previously used data on measuring green space use are self-reported or collected via social surveys, which not only include limited samples, but also are always subjective, costly, and laborious. This study integrates sensor and positioning technologies and measures the use of green space from an emerging big data perspective. The hourly real-time Tencent user density (RTUD) data from social media are used to analyze the time-spatial distribution of urban park users. RTUD data, park attributes, and surrounding landscape features are incorporated into ArcGIS for spatial analysis. A group of linear regression models is constructed to determine factors that may be associated with the user density of urban parks. The total accumulated number of observed users is 3.25 million in 686 urban parks of Shenzhen in two typical sunny days – a work day and a rest day. Without costly and laborious field investigation, based on RTUD data, we conduct a city wide analysis regarding all parks. The proposed method is proved to be suitable for measuring urban green space use at city-level or even large scale. Results show that park user density is relatively high in well developed areas and community parks. Park attributes and surrounding landscape features are significantly associated to park use. The findings of this study can help policy makers optimize the construction and maintenance of urban parks.