Representing Syntax and Composition with Geometric Transformations

TL;DR

This paper explores using geometric transformations to encode syntactic graphs in distributional semantic models, aiming to improve phrase composition while reducing model complexity and addressing data sparsity issues.

Contribution

It introduces a novel approach applying geometric transformations from knowledge graph models to syntactic graphs, enhancing syntactic relation encoding for semantic composition.

Findings

Geometric transformations effectively encode syntactic relations.

The approach reduces parameter count compared to traditional syntactically-aware DSMs.

Enhanced phrase representations via syntactic contextualization.

Abstract

The exploitation of syntactic graphs (SyGs) as a word's context has been shown to be beneficial for distributional semantic models (DSMs), both at the level of individual word representations and in deriving phrasal representations via composition. However, notwithstanding the potential performance benefit, the syntactically-aware DSMs proposed to date have huge numbers of parameters (compared to conventional DSMs) and suffer from data sparsity. Furthermore, the encoding of the SyG links (i.e., the syntactic relations) has been largely limited to linear maps. The knowledge graphs' literature, on the other hand, has proposed light-weight models employing different geometric transformations (GTs) to encode edges in a knowledge graph (KG). Our work explores the possibility of adopting this family of models to encode SyGs. Furthermore, we investigate which GT better encodes syntactic relations, so that these representations can be used to enhance phrase-level composition via syntactic contextualisation.