Keywords: semantic typology, parallel corpora, spatial relations, evaluation
TL;DR: Spatial relation markers are successfully extracted across a massively parallel corpus
Abstract: This paper presents a novel task, detecting Spatial Relation Markers (SRMs), like English 'on the boat', across languages, alongside a model, RUIMTE to approach this task with. Using a massively parallel corpus of Bible translations, the model is evaluated against existing
and baseline models on the basis of a novel evaluation set. The model presents high quality SRM extraction, as well as an accurate identification of situations in which no SRM is given in that language (‘zero marking’).
Submission Number: 212
Loading