Abstract
Distributional approaches of natural language semantics generally take a static approach, in which a semantic space is computed once for a full corpus in order to quantify term similarity. The current research proceeds towards a more dynamic approach, by measuring differences in semantic space within subsequent sections of a document
... read more