2019年度 人工知能学会全国大会（第33回）
2019年6月4日〜6月7日朱鷺メッセ 新潟コンベンションセンター
2019年度 人工知能学会全国大会（第33回）
2019年6月4日〜6月7日朱鷺メッセ 新潟コンベンションセンター
[2A4-E-2-03]Exploring Machine Learning Techniques for Irony Detection
〇Zheng Lin Chia1, Michal Ptaszynski1, Fumito Masui1(1. Kitami Institute of Technology)
Irony detection is considered a complex task in Natural Language Processing. This paper first introduce and cover the recently state of irony detection. Then we review and summarize previous related research on text-based irony detection. Finally we compare various classifiers including the proposed CNN model on three dataset of tweets, and analysis and discuss the results. We conclude that CNN is effective for irony detection under various situation with our model outperforming all the other classifiers.