Abstract: Under the dual drive of capital and scientific strength,shared bicycles suddenly break into our city and develop rapidly,becoming a part of the urban landscape and an indispensable way of transportation for citizens.However,in the same time,it also brings severe pressure and challenge to the urban traffic operation management and planning development.In this paper,the location data of ofo bicycles in September 20,2017 was obtained by means of Python program.Then,using the ArcGIS traditional analysis tools,the spatial-temporal feature of riding behavior for Xi'an users were analyzed,the hot spots of urban riding were extracted,and the user travel OD model was developed.The results indicated that riding lines had different spatial characteristics under the influence of urban rail transportation,urban industrial distribution,hospital and school distribution and time dimension.The hot spots derived by big data analysis collaborated with the results of comprehensive analysis of visual expression data sets as well as related city data sets makes a new way of thinking for the centralized planning of non-motorvehicle facilities and the optimization for urban spatial layout.

Key words: sharing-bikes' data, hot spot extraction, greedy algorithm, network analysis, planning bike lanes