To Cause Or Not To Cause: Cross-Lingual Semantic Matching for Paraphrase
Modelling

We presenta manual pilot study in cross-linguistic analysis at the
predicate-argument level. Looking at instances of a cross-POS translation
pair, we find that even though predicate-argument structure abstracts
from morphosyntactic language idiosyncrasies, there is still considerable
variation in the distribution of semantic material to predicates.  We
propose an algorithm for automatically identifying matching groups of
predicate-argument structures (frame paraphrases). The resulting data
allows an analysis of semantic differences e.g.\ in the expression of
different degrees of causality.