A Multi-task Ensemble Framework for Emotion, Sentiment and Intensity Prediction
This work addresses emotion and sentiment prediction problems for applications in social media and text analysis, but it is incremental as it builds on existing deep learning models and ensemble techniques.
The paper tackles emotion and sentiment analysis across multiple granularities and domains by proposing a multi-task ensemble framework that combines CNN, LSTM, GRU, and hand-crafted features, resulting in an average performance improvement of 2-3 points over single-task systems.
In this paper, through multi-task ensemble framework we address three problems of emotion and sentiment analysis i.e. "emotion classification & intensity", "valence, arousal & dominance for emotion" and "valence & arousal} for sentiment". The underlying problems cover two granularities (i.e. coarse-grained and fine-grained) and a diverse range of domains (i.e. tweets, Facebook posts, news headlines, blogs, letters etc.). The ensemble model aims to leverage the learned representations of three deep learning models (i.e. CNN, LSTM and GRU) and a hand-crafted feature representation for the predictions. Experimental results on the benchmark datasets show the efficacy of our proposed multi-task ensemble frameworks. We obtain the performance improvement of 2-3 points on an average over single-task systems for most of the problems and domains.