Multi-layered Annotation of Non-textual Data for Spatial Information

Spatial and spatio-temporal information is often carried by non-textual data such as maps, diagrams, tables, or pictures, both still and moving, either embedded in a text or standalone. The annotation of nontextual data raises the following questions: (i) what are the markables and how should they be coded? (ii) how should relevant information be inferred which is implicit in the data? We answer these questions with a multilayered approach.