Experimenting with Transitive Verbs in a DisCoCat

TL;DR

This paper explores the use of transitive verbs as matrices within the DisCoCat framework, evaluating different construction methods through a disambiguation task to enhance understanding of compositional semantics.

Contribution

It introduces three new methods for constructing transitive verb matrices in DisCoCat and evaluates their effectiveness in a disambiguation task.

Findings

All methods improved disambiguation accuracy

One method outperformed others in specific contexts

The study advances practical applications of DisCoCat models

Abstract

Formal and distributional semantic models offer complementary benefits in modeling meaning. The categorical compositional distributional (DisCoCat) model of meaning of Coecke et al. (arXiv:1003.4394v1 [cs.CL]) combines aspected of both to provide a general framework in which meanings of words, obtained distributionally, are composed using methods from the logical setting to form sentence meaning. Concrete consequences of this general abstract setting and applications to empirical data are under active study (Grefenstette et al., arxiv:1101.0309; Grefenstette and Sadrzadeh, arXiv:1106.4058v1 [cs.CL]). . In this paper, we extend this study by examining transitive verbs, represented as matrices in a DisCoCat. We discuss three ways of constructing such matrices, and evaluate each method in a disambiguation task developed by Grefenstette and Sadrzadeh (arXiv:1106.4058v1 [cs.CL]).