Mittens: An Extension of GloVe for Learning Domain-Specialized Representations

TL;DR

This paper introduces Mittens, an extension of GloVe that adapts general-purpose word embeddings to specialized domains, resulting in faster learning and improved task performance.

Contribution

Mittens is a simple method that fine-tunes GloVe embeddings with domain-specific data to enhance their usefulness in specialized tasks.

Findings

Faster learning with domain-adapted embeddings

Improved performance on domain-specific tasks

Effective extension of GloVe for specialized domains

Abstract

We present a simple extension of the GloVe representation learning model that begins with general-purpose representations and updates them based on data from a specialized domain. We show that the resulting representations can lead to faster learning and better results on a variety of tasks.