A Novel Spatial-Temporal Fusion Framework Based on Object Trajectories

With the development of information technology, the traces of residents could be caught by sensors all over the city, from which to identify specific people such as suspects, has become the focus of public security departments in recent years. Although there are some advantages of the mass of Spatio-temporal records, the activity records contain both temporal and spatial context is high-order and sparse. Therefore, it is difficult to identify users, especially for those small groups of individuals. In this paper, we propose a framework to identify users by fuzzed Spatio-temporal traces, in which we first extract the features from several perspectives and generate the TSGH (Time-slice Geohash) model and STPS (Spatio-temporal Proportion and Similarity) model. All these features are combined with classical classification models to distinguish users. We set up a real-life dataset and evaluate performance effectiveness with precision and recall measures. The final results show that the fuzzy operations could describe the characteristic of traces better and improve the prediction performance.