A Human Mobility Prediction Scheme By Using A Hierarchical Interest Model

摘要

We propose a scheme to predict human mobility in this technical report. First, a hierarchical interest model is introduced to organize the semantic category of location in the human mobility logs and to represent the personalized mobility patterns of a single person. Then, by combining the interest models of many different people, a 3-dimensional tensor with the features of person, time, and the semantic category of location is constructed. Tensor factorization is utilized to reveal people's mobility interest on different kinds of locations. Finally, personalized interest models are recovered from the cumulative tensor and are used to predict human mobility in a person-by-person way. Extensive evaluation results based on a large scale dataset of real check-in records have validated that our proposal achieves better recall, precision, and F-Score in human mobility prediction as compared to the state-of-art approach.