Real-time trajectory privacy protection based on improved differential privacy method and deep learning model

Journal of Cloud Computing

Advances, Systems and Applications

Table 1 Dataset description

Dataset	Acquisition	Frequency	Time span	Application	Dimension
Yonsei	Site sensor	1 h	364 day	Monitored track	[day, timestep, nodes, 1]
Citi Bike	Loop detector	1 h	364 day	Shared bicycle	[day, timestep, nodes, 1]

(1) The Yonsei dataset comprises pedestrian trajectory data generated by nine graduate students over a two-month period in 2011 using a cell phone location service application called SmartDC. The data were collected in Yonsei University in Seoul, South Korea. The data size is [33203, 4], and the data features contain person_id, time, latitude, and longitude
(2) The Citi Bike dataset comprises bike-sharing trajectory data across more than 600 stations in Toronto, Canada, in 2016. We used data features including time, latitude, and longitude