Semantic Time-Geographic Modelling and Analysis

Abstract. Time geography is a powerful framework that provides a set of well-defined space-time entities and relationships for representing human activity-travel behaviors under various space-time constraints. The classical time-geographic framework focuses on only the geometric aspect of activity-travel behaviors in the space and time dimensions but ignores rich semantic contexts of activity-travel behaviors (e.g., activity goals, transport modes, and individual emotions and feelings). This study introduces a novel time-geographic framework by explicitly incorporating semantic contexts of activity-travel behaviors. The introduced framework provides rigorous definitions for full spectrum of semantically rich space-time entities (e.g., semantic prisms and lifelines) and relationships (e.g., semantic space-time intersection relationships). The introduced framework is implemented by using two types of database approaches: relational and graph databases. Experimental results demonstrate the capabilities of the introduced semantic time-geographic framework for representing both geometric and semantic perspectives of human activity-travel behaviors. The graph database approach can achieve various queries on a large number of semantic time-geographic entities within reasonable computational times.