Analysis of the Polysemy Exhibited by Spatial Prepositions

In previous work exploring the semantics of spatial prepositions in grounded settings, we conducted a study to gather annotations related to spatial language in 3D virtual environments (https://doi.org/10.5518/764). Using this collected data we initially considered a semantic model based on Prototype Theory and introduced a method for learning its parameters from data which we discuss in http://eprints.whiterose.ac.uk/156866/. However, though there is much to suggest that spatial prepositions exhibit polysemy, each term was treated as exhibiting a single sense. The ability for terms to represent distinct but related meanings is unexplored in the work on grounded semantics and referring expressions, where even homonymy is rarely considered. Using the previously collected data we have examined how the polysemy exhibited by spatial prepositions may be incorporated into semantic models for situated dialogue and we have tested the approaches in a reference task. The current dataset provides an archive of the associated code and results of this analysis.