Multiclass mood classification on Twitter using lexicon dictionary and machine learning algorithms

Sentiment Analysis (SA), an application of Natural Language processing (NLP), has evolved a lot over the past decade. It is also known as opinion mining, mood extraction and emotion analysis. Social Media provides a good platform for people to share their thoughts, feelings and views. Social Media like Twitter is flooded everyday with tweets from users. These tweets can be effectively used for extracting mood of a person. Extraction of mood from texts is a challenging task as it involves understanding the underlying semantic. In this paper a new approach is been proposed that uses lexicon database to assign each word in a text a value called as `Impact Factor'. The Impact Factor is nothing but how a single word is affecting the whole sentence in which it is used. Every word in a sentence has its own Impact Factor and it tries to influence the overall semantic of the sentence. Higher the value of Impact Factor of a word in the sentence, the more influential it is. The approach proposed in this paper makes use of lexicon based approach as well as machine based learning. It uses AFINN lexicon database to assign Impact Factor to words and Support Vector Machine (SVM), k-Nearest Neighbors (KNN) and Naive Bayesian (NB) machine learning algorithms for training and testing the model.