NSEmo at EmoInt-2017: An Ensemble to Predict Emotion Intensity in Tweets

Abstract

In this paper, we describe a method to predict emotion intensity in tweets. Our approach is an ensemble of three regression methods. The first method uses contentbased features (hashtags, emoticons, elongated words, etc.). The second method considers word n-grams and character ngrams for training. The final method uses lexicons, word embeddings, word ngrams, character n-grams for training the model. An ensemble of these three methods gives better performance than individual methods. We applied our method on WASSA emotion dataset. Achieved results are as follows: average Pearson correlation is 0.706, average Spearman correlation is 0.696, average Pearson correlation for gold scores in range 0.5 to 1 is 0.539, and average Spearman correlation for gold scores in range 0.5 to 1 is 0.514.

THU_NGN at SemEval-2018 Task 1: Fine-grained Tweet Sentiment Intensity Analysis with Attention CNN-LSTM

“…Different approaches have been proposed to detect the tweet emotion intensity in the EmoInt-2017 shared task (Mohammad and Bravo-Marquez, 2017). For example, Madisetty et al (2017) proposed an ensemble model based on SVR. Goel et al (2017) and Koper et al (2017) applied CNN-LSTM architecture to this task.…”

“…However, these method can't utilize the contextual information from texts. Supervised methods are mainly based on SVR (Madisetty and Desarkar, 2017), linear regression (John and Vechtomova, 2017) and neural networks (Goel et al, 2017;Köper et al, 2017). Usually neural network-based methods outperform SVR and linear regression-based methods siginificantly.…”