Study On Human Mobility Pattern

The exploration on human behavior, especially mining and modeling temporal and spatial regularities, is a common focus in nature,economics and sociology. Analysis on the spatial regularities of human mobility helps in explaining many phenomena in disease propagation、traffic controlling、abnormal detection, migration of population etc. Mobility is a central aspect of our life; the locations we visit reflect our tastes and lifestyle and shape our social relationships. The human mobility pattern in urban ordinary life is influenced by various factors.To understanding the features of human mobility in urban, a dataset of location information reported by mobile phones is utilized to analyze the spatial mobility pattern of check-ins and a series of related predictions are proposed under the analysis results.The mobility pattern analysis contains the factor analysis on human mobility predictability and the impacts of factors including travel distance、social relationship、age and gender.It’s found that the check-ins show certain memory effect. The number of visited locations and the visiting pattern to the most frequent visited location have more significant influence on the predictability and regularity of check-ins, meanwhile the impacts of radius of gyration and the average jump distance are obviously unremarkable.The patterns of human regular and long check-ins dividing by the jump distance are different in POI of visited locations and their spatial and temporal features.Social relationships have impacts on human check-ins. It’s found that the number of common friends and the strength of friendship shows greater influence on friends’ check-ins, while the impacts of direction of relationship are unremarkable. Differences in gender and age make distinctions of check-ins among visited locations,mobility behavior and social relationships.With all results above, the prediction of user’s gender and age and the trajectory prediction have been proceeded which has been gained the accuracy of 75% of gender prediction,74% of age prediction and 56% of trajectory prediction.