Extreme Learning Machine for Multi-class Sentiment Classification of Tweets

The increasing popularity of social media in recent years has created new opportunities to study and evaluate public opinions and sentiments for use in marketing and social behavioural studies. However, binary classification into positive and negative sentiments may not reveal too much information about a product or service. This research paper explores the multi-class sentiment classification using machine learning methods. Three machine learning methods are investigated in this paper to examine their respective performance in multi-class sentiment classification of tweets. Experimental results show that Extreme Learning Machine (ELM) achieves better performance than other machine learning methods.