IIIDYT at SemEval-2018 Task 3: Irony detection in English tweets

TL;DR

This paper presents a bidirectional LSTM-based system for irony detection in English tweets, achieving better validation results but limited generalization, highlighting the need for more pre-training on small datasets.

Contribution

Introduces a representation learning approach using multi-layered bidirectional LSTM without external features for irony detection in tweets.

Findings

Outperforms baseline on validation set

Limited generalization on test set

Suggests pre-training could improve results

Abstract

In this paper we introduce our system for the task of Irony detection in English tweets, a part of SemEval 2018. We propose representation learning approach that relies on a multi-layered bidirectional LSTM, without using external features that provide additional semantic information. Although our model is able to outperform the baseline in the validation set, our results show limited generalization power over the test set. Given the limited size of the dataset, we think the usage of more pre-training schemes would greatly improve the obtained results.