Abstract
Spatial Reasoning from language is essential for natural languageunderstanding. Supporting it requires a representation scheme that can capturespatial phenomena encountered in language as well as in images and videos.Existing spatial representations are not sufficient for describing spatialconfigurations used in complex tasks. This paper extends the capabilities ofexisting spatial representation languages and increases coverage of thesemantic aspects that are needed to ground the spatial meaning of naturallanguage text in the world. Our spatial relation language is able to representa large, comprehensive set of spatial concepts crucial for reasoning and isdesigned to support the composition of static and dynamic spatialconfigurations. We integrate this language with the Abstract MeaningRepresentation(AMR) annotation schema and present a corpus annotated by thisextended AMR. To exhibit the applicability of our representation scheme, weannotate text taken from diverse datasets and show how we extend thecapabilities of existing spatial representation languages with the fine-graineddecomposition of semantics and blend it seamlessly with AMRs of sentences anddiscourse representations as a whole.