Analysis of Machine-Learning Techniques for Emotion Detection from Text

Emotion detection plays a very pivotal role in decision-making in business, in our day-to-day life and interpersonal relationships. Hence, identifying the emotion from text, speech, and videos has been a research area from long time. Therefore, there are several research happened with new advancements. This paper presents the analysis of machine-learning-based approaches for classifying the emotion from text. There are various machine-learning algorithms used, and results are compared for emotion classification model from text data. For this purpose, ISEAR (International Survey on Emotion Antecedents and Reactions) dataset is used and achieved close to 60% of accuracy as overall in validation dataset. Objective of this paper is to compare the performance of various machine-learning algorithms on text-based emotion detection.