VCDM: Leveraging Variational Bi-encoding and Deep Contextualized Word Representations for Improved Definition Modeling

TL;DR

This paper introduces VCDM, a generative model using variational inference and contextualized embeddings to improve word and phrase definition modeling, achieving state-of-the-art results on multiple benchmarks.

Contribution

The paper presents a novel generative approach with explicit latent variables for definition modeling, leveraging contextualized embeddings and introducing new datasets.

Findings

VCDM outperforms existing models on four benchmarks.

The model achieves higher accuracy in automatic and human evaluations.

New datasets 'Cambridge' and 'Robert' enhance empirical analysis.

Abstract

In this paper, we tackle the task of definition modeling, where the goal is to learn to generate definitions of words and phrases. Existing approaches for this task are discriminative, combining distributional and lexical semantics in an implicit rather than direct way. To tackle this issue we propose a generative model for the task, introducing a continuous latent variable to explicitly model the underlying relationship between a phrase used within a context and its definition. We rely on variational inference for estimation and leverage contextualized word embeddings for improved performance. Our approach is evaluated on four existing challenging benchmarks with the addition of two new datasets, "Cambridge" and the first non-English corpus "Robert", which we release to complement our empirical study. Our Variational Contextual Definition Modeler (VCDM) achieves state-of-the-art performance in terms of automatic and human evaluation metrics, demonstrating the effectiveness of our approach.