Construction of semantic annotation framework for stacked trajectory model

The trajectory data contains not only interesting facts about individual trajectory levels, but also levels of trajectory sets that display certain features. These features can come from spatial and temporal dependencies, or from attribute characteristics under spatiotemporal dynamic conditions. An important goal of seeking a better understanding of trajectory behavior research is, for example, the ability to express and predict how trajectory data responds to changes in their environment and how these responses are related in time, space, and attributes. However, this is not a straightforward task, as it involves not only the decision process, but also the semantic constraints from the context in abstraction process of animation. We build a computational framework for semantic annotation to facilitate analysis and discovery of characterization in trajectory behavior, and to evaluate our approach using case studies of open-pit mine truck datasets.