Yuan at SemEval-2018 Task 1: Tweets Emotion Intensity Prediction using Ensemble Recurrent Neural Network

Proceedings of the 12th International Workshop on Semantic Evaluation (SemEval-2018), pages 205–209 New Orleans, Louisiana, June 5–6, 2018. ©2018 Association for Computational Linguistics Yuan at SemEval-2018 Task 1: Tweets Emotion Intensity Prediction using Ensemble Recurrent Neural Network Min Wang, Xiaobing Zhou* School of Information Science and Engineering Yunnan University, Yunnan, P.R. China *Corresponding author, zhouxb.cn@gmail.com Abstract This paper describes the performing system for SemEval-2018 Task 1 subtask 3 - Given a tweet, determine the intensity of sentiment or valence (V) that best represents the mental state of the tweeter—a real-valued score between 0 (most negative) and 1 (most positive). The proposed system gets features in tweets from the existing emotional dictionary and represents the word using word emb- edding, then utilizes the joint repre- sentations as the inputs of the bidire- ctional long short-term memory (BiL- STM) to learn and get the regression result. To boost performance we ensem- ble several BiLSTMs together. We ranked 6th in subtask 3 among all teams. Our approach achieves the Pearson(All instances) score 0.836 and Pearson(gold in 0.5-1) score 0.667, we outperform the baseline model of this task by 25.1% and 21.8% of Pearson(All instances) and Pearson(gold in 0.5-1) scores respectively. 1 Introduction Sentiment analysis (SA) is a field of knowledge which deals with the analysis of people’s opinions, sentiments, evaluations, appraisals, attitudes and emotions towards particular entities (Liu, 2012). EmoInt (Mohammad and Bravo- Marquez, 2017) is a shared task hosted by WASSA 2017, aiming to predict the emotion intensity in tweets. SemEval 2018 Task 1 subtask 3 (Mohammad et al, 2018) is similar to EmoInt, however the goal of subtask 3 is to detect valenc- e or sentiment intensity, in which scores are floating point values between 0 and 1, representing low and high intensities of the emotion being expressed, respectively. Obviously we don’t know in advance whether twitter’s emotional intensity is positive or negative, but in EmoInt task we can determine whether twitter emotions are positive or negative based on one of four datasets: anger, fearness, joy, sadness. This is still a challenging task and remains active areas of research. These setbacks are: extensive usage of hashtags, slang, abbreviations, and emoticons. And tweets are usually typed on mobile devices like mobile phone, laptop or iPad which can result in a substantial amount of typos. Existing methods for modeling emotion intensity rely vastly on manually constructed lexicons, which contain information about intensity weights for each available word (Mohammad and Bravo-Marquez, 2017a; Neviarouskaya et al., 2007). The intensity for the whole tweet can be deduced by combining individual scores of words, which is easy and ignores the word order compositionality of the language. Building such lexicons is a labour- intensive procedure. We can learn from these models the skills of combining feature extraction and classification or regression stages given a sufficient amount of training data. Some deep learning methods are used to process the same question. Deep neural archit- ectures for emotion intensity prediction in tweets (Goel et al., 2017) and character- and word-level …

Content truncated to 500 words