Sentiment Analysis on Social Network: Using Emoticon Characteristics for Twitter Polarity Classification

Computational Linguistics and Chinese Language Processing Vol. 23, No. 1, June 2018, pp. 1-18 1  The Association for Computational Linguistics and Chinese Language Processing Sentiment Analysis on Social Network: Using Emoticon Characteristics for Twitter Polarity Classification Chia-Ping Chen, Tzu-Hsuan Tseng and Tzu-Hsuan Yang Abstract In this paper, we describe a sentiment analysis system implemented for the semantic-evaluation task of message polarity classification for English on Twitter. Our system contains modules of data pre-processing, word embedding, and sentiment classification. In order to decrease the data complexity and increase the coverage of the word vector model for better learning, we perform a series of data pre-processing tasks, including emoticon normalization, specific suffix splitting, and hashtag segmentation. In word embedding, we utilize the pre-trained word vector provided by GloVe. We believe that emojis in tweets are important characteristics for Twitter sentiment classification, but most pre-trained sets of word vectors contain few or no emoji representations. Thus, we propose embedding emojis into the vector space by neural network models. We train the emoji vector with relevant words that contain descriptions and contexts of emojis. The models of long short-term memory (LSTM) and convolutional neural network (CNN) are used as our sentiment classifiers. The proposed emoji embedding is evaluated on the SemEval 2017 tasks. Using emoji embedding, we achieved recall rates of 0.652 with the LSTM classifier and 0.640 with the CNN classifier. Keywords: Sentiment Analysis, Polarity Classification, Machine Learning, Neural Network, Word Embedding. 1. Introduction There has been huge growth in the use of social networks, such as Twitter, in recent years. Many messages are created every day, including various topics, users’ comments and views, or current emotions. Sentiment analysis, which predicts the polarity of a message, is one of the research directions on Twitter. A message on Twitter is called a tweet and is allowed to be 140 characters or less. Tweets are highly colloquial. Due to the length constraint, a tweet often  Department of Computer Science and Engineering, National Sun Yat-sen University, Taiwan E-mail: cpchen@cse.nsysu.edu.tw; fb74123698@gmail.com; kr60903@gmail.com 2 Chia-Ping Chen et al. contains unofficial abbreviations, as well as emoticons and emojis. Figure 1 shows examples of tweets. Figure 1. Examples of tweets. Tweets tend to have informal words and syntax. In the above examples, we can see that emojis are used frequently in Tweets. Some emojis (like ) can be considered the natural evolution of emoticons, such as :-) and :D. In addition to facial expressions, emojis can be used for food, flags, animals, etc. Unofficial abbreviations and emojis without corresponding word vectors in tweets can make the sentiment classification task difficult. In this work, we find sentiment features for these unorthodox tokens to get better results in sentiment classification. Artificial neural networks for machine learning are mathematical models inspired by biological neural systems. Deep learning, which is neural network models based on deep neural networks, has been a very successful method and achieves state-of-the-art performance in many tasks, such as NIST handwritten digit recognition (LeCun, Bottou, Bengio, & Haffner, …

Content truncated to 500 words