Detecting Unseen Multiword Expressions in American Sign Language

TL;DR

This paper explores using GloVe word embeddings to detect multiword expressions in American Sign Language, demonstrating that embeddings can identify non-compositional phrases with reasonable accuracy.

Contribution

It introduces a system leveraging word embeddings for multiword expression detection in ASL, a novel application in sign language translation.

Findings

Word embeddings can detect non-compositionality with decent accuracy

GloVe embeddings contain useful data for multiword expression prediction

The approach shows promise for improving sign language translation systems

Abstract

Multiword expressions present unique challenges in many translation tasks. In an attempt to ultimately apply a multiword expression detection system to the translation of American Sign Language, we built and tested two systems that apply word embeddings from GloVe to determine whether or not the word embeddings of lexemes can be used to predict whether or not those lexemes compose a multiword expression. It became apparent that word embeddings carry data that can detect non-compositionality with decent accuracy.