A Unified Model for Reverse Dictionary and Definition Modelling

TL;DR

This paper introduces a dual-way neural model that simultaneously retrieves words from definitions and generates definitions from words, demonstrating strong performance and human preference without extra resources.

Contribution

It presents a unified multi-task neural framework for reverse dictionary and definition modeling, effectively handling unknown words and improving over previous benchmarks.

Findings

Achieves promising automatic scores on benchmarks.

Human annotators prefer the model's outputs.

Multiple objectives enhance learning.

Abstract

We build a dual-way neural dictionary to retrieve words given definitions, and produce definitions for queried words. The model learns the two tasks simultaneously and handles unknown words via embeddings. It casts a word or a definition to the same representation space through a shared layer, then generates the other form in a multi-task fashion. Our method achieves promising automatic scores on previous benchmarks without extra resources. Human annotators prefer the model's outputs in both reference-less and reference-based evaluation, indicating its practicality. Analysis suggests that multiple objectives benefit learning.