One Representation per Word - Does it make Sense for Composition?
Abstract
In this paper, we investigate whether an a priori disambiguation of word senses is strictly necessary or whether the meaning of a word in context can be disambiguated through composition alone. We evaluate the performance of off-the-shelf single-vector and multi-sense vector models on a benchmark phrase similarity task and a novel task for word-sense discrimination. We find that single-sense vector models perform as well or better than multi-sense vector models despite arguably less clean elementary representations. Our findings furthermore show that simple composition functions such as pointwise addition are able to recover sense specific information from a single-sense vector model remarkably well.
Keywords
Cite
@article{arxiv.1702.06696,
title = {One Representation per Word - Does it make Sense for Composition?},
author = {Thomas Kober and Julie Weeds and John Wilkie and Jeremy Reffin and David Weir},
journal= {arXiv preprint arXiv:1702.06696},
year = {2017}
}
Comments
to appear at the EACL 2017 workshop on Sense, Concept and Entity Representations and their Applications