Multiclass Emotion Analysis Using NLP

Rohan Madhani, Sagar Makwana, Viral Lakhani, Alabh Mehta, Sindhu Nair

Abstract: In the present scenario, sentiment analysis has become a popular topic in the field of Machine Learning (ML) and Natural Language Processing (NLP). Sentiment analysis is the systematic process of determining the sentimental tone in a array of words. It helps to understand the emotion, attitudes, and opinions expressed in the sentence. Machine learning techniques are widely used in determining the emotions from texts due to their precise prediction. Various classifiers can be used for performing sentiment analysis which may provide different accuracy. This paper documents a comparative study of three machine learning classifiers namely, Support Vector Machine (SVM), Recurrent Neural Network (RNN) - Long Short Term Memory (LSTM) and Naive Bayes for performing seven class sentiment analysis. The emotions which were considered for this study were: joy, sadness, anger, shame, guilt, disgust and fear. For analyzing the performance of these models precision, recall and F1 score were used. From the result, we come to know that the performance of a Recurrent Neural Network was much better than other classifiers

Keywords: Sentiment Analysis, Machine Learning, Support Vector Machine, Recurrent Neural Network, Naive Bayes

How to Cite?: Rohan Madhani, Sagar Makwana, Viral Lakhani, Alabh Mehta, Sindhu Nair, "Multiclass Emotion Analysis Using NLP", Volume 7 Issue 11, November 2018, International Journal of Science and Research (IJSR), Pages: 336-338, https://www.ijsr.net/getabstract.php?paperid=ART20192599, DOI: https://dx.doi.org/10.21275/ART20192599