FrameSTEP: A framework for annotating semantic trajectories based on episodes

•A high-level data model for representing trajectory episodes and context is proposed.•Spatial and temporal extents as well as granularity levels can be easily expressed.•Quantitative and qualitative semantic descriptions can be used to annotate episodes.•A framework for trajectory annotation shows how the model can be instantiated.•A spatial annotation method that uses OpenStreetMap data is shown. We are witnessing an increasing usage of location data by a variety of applications. Consequently, information systems are required to deal with large datasets containing raw data to build high level abstractions. Semantic Web technologies offer powerful representation tools for pervasive applications. The convergence of location-based services and Semantic Web standards allows an easier interlinking and annotation of trajectories. However, due to the wide range of requirements on modeling mobile object trajectories, it is important to define a high-level data model for representing trajectory episodes and contextual elements with multiple levels of granularity and different options to represent spatial and temporal extents, as well as to express quantitative and qualitative semantic descriptions. In this article, we focus on modeling mobile object trajectories in the context of Semantic Web. First, we introduce a new version of the Semantic Trajectory Episodes (STEP) ontology to represent generic spatiotemporal episodes. Then, we present FrameSTEP as a new framework for annotating semantic trajectories based on episodes. As a result, we combine our ontology, which can represent spatiotemporal phenomena at different levels of granularity, with annotation algorithms, which allow to create instances of our model. The proposed spatial annotation algorithm explores the Linked Open Data cloud and OpenStreetMap tags to find relevant types of spatial features in order to describe the environment where the trajectory took place. Our framework can guide the development of future expert systems in trajectory analysis. It enables reasoning about knowledge gathered from large trajectory data and linked datasets in order to create several intelligent services.