A Multivariate Model for Representing Semantic Non-compositionality

TL;DR

This paper introduces a multivariate model that effectively identifies semantically non-compositional phrases by integrating multiple characteristics, outperforming previous models that focus on single features.

Contribution

The paper presents a novel multivariate model that combines various characteristics to improve the identification of non-compositional phrases in NLP.

Findings

Model outperforms existing approaches

Integrates multiple characteristics for better accuracy

Highlights importance of combined features in semantic analysis

Abstract

Semantically non-compositional phrases constitute an intriguing research topic in Natural Language Processing. Semantic non-compositionality --the situation when the meaning of a phrase cannot be derived from the meaning of its components, is the main characteristic of such phrases, however, they bear other characteristics such as high statistical association and non-substitutability. In this work, we present a model for identifying non-compositional phrases that takes into account all of these characteristics. We show that the presented model remarkably outperforms the existing models of identifying non-compositional phrases that mostly focus only on one of these characteristics.