Spatiotemporal Analysis To Observe Activities Behavior And City Rhythm By Using Social Media Big Data:A Case Study Of Shanghai,China

Location based social network(LBSN)is prevailing rapidly in China with increase and adoption of smart devices which provide a wide range of opportunities to observe and analyze the human activities behavior towards the use of LBSN.In LBSN,users socialize with each other by sharing their current location(also referred as "geolocation")in the form of a tweet(also referred as "check-in"),which contains(but not limited to)text,audio,video etc.,and allows users to disclose the places they visit and activities(i.e.eating,living,working and playing)performed.Understanding this human activities behavior in space and time on LBSN datasets also called "activities behavior" can archive the day-to-day activity patterns,usage behaviors towards social media and presents spatiotemporal evidence of user’s daily routines.LBSN datasets do not fully reveal the specific human activities behavior due to certain limitations(i.e.,biases in gender,low sampling rate and biases in location type).Therefore due to these limitations,LBSN data is only considered to be a supplement to traditional methods(i.e.,survey,census)for observing human activities behavior.This phenomenon of sharing location,activities,and time by users on LBSN has encouraged this research to observe and analyze the human activities behavior and the city rhythm(also referred as "movement of the city based on activities").The primary objective of this research is to observe the users behavior towards different activities in space and time and analyze the principle orientation,direction,spatiotemporal trends,and the distribution differences of activities at the city and district level by using LBSN data.Kernel density estimation(KDE)is utilized to explore the spatiotemporal distribution of activities,geographically and weighted regression(GWR)method was applied to observe the relationship between activities and districts with a focus on gender,gravity center(hereby referred as "activities weighted mean center"(AWMC))and standard deviational ellipse(SDE)analysis is used to observe and analyze the orientation,direction,spatiotemporal expansion trends of activities and the city rhythm in the study area.The results of this study show that LBSN can be considered as a reliable source of information to observe human activities behavior within a city in space and time.Furthermore,it shows that female users are more likely to use social media use as compared to male.However,the activities behavior patterns for male users is quite different during weekdays and weekend as compared to female users.In summary,the results of SDE at the district level can better reflect the change trajectory of spatiotemporal dynamics of activities and its fine spatial structure than at the whole city level.