UTFPR at IEST 2018: Exploring Character-to-Word Composition for Emotion Analysis

Abstract

We introduce the UTFPR system for the Implicit Emotions Shared Task of 2018: A compositional character-to-word recurrent neural network that does not exploit heavy and/or hard-to-obtain resources. We find that our approach can outperform multiple baselines, and offers an elegant and effective solution to the problem of orthographic variance in tweets.

Tasks

Reproductions