NTUA-SLP at SemEval-2018 Task 1: Predicting Affective Content in Tweets with Deep Attentive RNNs and Transfer Learning
This work addresses emotion analysis in social media for NLP applications, with incremental improvements in model interpretability and performance.
The paper tackled predicting affective content in tweets using deep attentive RNNs and transfer learning, achieving first place in multi-label emotion classification and second in emotion intensity regression at SemEval-2018.
In this paper we present deep-learning models that submitted to the SemEval-2018 Task~1 competition: "Affect in Tweets". We participated in all subtasks for English tweets. We propose a Bi-LSTM architecture equipped with a multi-layer self attention mechanism. The attention mechanism improves the model performance and allows us to identify salient words in tweets, as well as gain insight into the models making them more interpretable. Our model utilizes a set of word2vec word embeddings trained on a large collection of 550 million Twitter messages, augmented by a set of word affective features. Due to the limited amount of task-specific training data, we opted for a transfer learning approach by pretraining the Bi-LSTMs on the dataset of Semeval 2017, Task 4A. The proposed approach ranked 1st in Subtask E "Multi-Label Emotion Classification", 2nd in Subtask A "Emotion Intensity Regression" and achieved competitive results in other subtasks.