Perceiving the emotions from the short-text using machine learning classifiers

In human-human communication, emotions are extremely important. Emotions determine a person’s willingness to act, persevere, smack, avoid danger, and be aware of others. Actions, text, facial expressions, and body language can all be used to portray emotions. Because of its impact on inter-personal communication, text emotions have become a prominent focus of research. It is the domain of natural language processing to identify the emotions from the text, using the sentimental analysis concept. In this paper, the emotions anger, sadness, surprise, happy, love and fear are detected from the text. The model is trained using the machine learning classifiers Support Vector Machine, K Nearest Neighbour, Logistic Regression and Random Forest. The most appropriate model is identified by testing using the metrics Precision, F1-score, Recall, accuracy, classification report and confusion matrix.