Zewen at SemEval-2018 Task 1: An Ensemble Model for Affect Prediction in Tweets

Proceedings of the 12th International Workshop on Semantic Evaluation (SemEval-2018), pages 313–318 New Orleans, Louisiana, June 5–6, 2018. ©2018 Association for Computational Linguistics Zewen at SemEval-2018 Task 1: An Ensemble Model for Affect Prediction in Tweets Zewen Chi, Heyan Huang, Jiangui Chen, Hao Wu, Ran Wei School of Computer Science, Beijing Institute of Technology, Beijing, China czwin32768@gmail.com, hhy63@bit.edu.cn, chenjiangui@outlook.com, wuhao123@bit.edu.cn, weiranbit@163.com Abstract This paper presents a method for Affect in Tweets, which is the task to automatically de- termine the intensity of emotions and inten- sity of sentiment of tweets. The term affect refers to emotion-related categories such as anger, fear, etc. Intensity of emotions need to be quantified into a real valued score in [0, 1]. We propose an ensemble system including four different deep learning methods which are CNN, Bidirectional LSTM (BLSTM), LSTM- CNN and a CNN-based Attention model (CA). Our system gets an average Pearson correla- tion score of 0.682 in the subtask EI-reg and an average Pearson correlation score of 0.784 in subtask V-reg, which ranks 19th among 48 systems in EI-reg and 17th among 38 systems in V-reg. 1 Introduction Affect determination is a significant part of nature language processing. Especially, affect in tweets becomes a focus in recent years. Sentiment Anal- ysis in Twitter, which is a task of SemEval, was firstly proposed in 2013 and not replaced until 2018. In SemEval 2018, the task Affect in Tweets (AIT) (Mohammad et al., 2018) was proposed and the objective is to automatically determine the in- tensity of emotions (E) and intensity of sentiment (aka valence V) of tweets. In this paper, we focus on two subtasks: • EI-reg (emotion intensity regression) – Given a tweet and an emotion E, determine the in- tensity of E that best represents the mental state of the tweeter – a real-valued score be- tween 0 (least E) and 1 (most E) • V-reg (sentiment intensity regression) – Given a tweet, determine the intensity of sen- timent or valence (V) that best represents the mental state of the tweeter – a real-valued score between 0 (most negative) and 1 (most positive) Before 2016, most systems use Support Vector Machine (SVM), Naive Bayes, maximum entropy and linear regression (Nakov et al., 2013; Rosen- thal et al., 2014, 2015). In SemEval 2014, deep learning methods started to appear and a team using them won the second place. Since 2015, more and more teams who were rank at the top used deep learning methods and now deep learn- ing methods including CNN and LSTM networks become really popular (Nakov et al., 2016; Rosen- thal et al., 2017). The system described in this paper is an en- semble of four different DNN methods including CNN, Bidirectional LSTM (Bi-LSTM), LSTM- CNN and a CNN-based Attention model (CA). In these methods, words in tweets are firstly mapped to word vectors. After intensity scores are calcu- lated by these models, we use a logistic regression and finally give the scores. The rest of the paper is organized …

Content truncated to 500 words