ISO-Space: Annotating Static and Dynamic Spatial Information

Abstract

An understanding of spatial information in natural language is necessary for many computational linguistics and artificial intelligence applications. In this chapter, we describe an annotation scheme for the markup of spatial relations, both static and dynamic, as expressed in text and other media. The desiderata for such a specification language are presented along with what representational mechanisms are required for such a specification to be successful. We review the annotation development process, and the adoption of the initial specification ISOspace, as an ISO standard, renamed ISOspace. We conclude with a discussion of the use of ISOspace in the context of the shared task SpaceEval 2015. The various ways in which natural languages make reference to spatial information has proved a challenge to both theoretical modeling as well as computational analysis. This is due largely to the diverse and heterogeneous strategies used to encode "spatial awareness" in text, ranging from space-denoting lexemes (motion verbs such as