Author : Anjali Deshpande and Prof. Ratnamala Paswan

The analysis of social media posts is extremely challenging as it concerns the detection of user communities. As emotions play a pivotal role in human interaction, the capability to detect them via analysing social media posts has various applications such as detecting psychological disorders in individuals or quantitatively detecting the public mood of a community. Previous studies on emotion classification made use of lexicons and bag-of-words classifiers. However, the existing work of emotion recognition on Twitter was carried out with the help of deep learning techniques on static Twitter data by taking into account only the hashtags present. The proposed method tries to increase the overall accuracy of emotion recognition via machine learning algorithms on real-time streaming data fetched from Twitter. The overall aim is to accurately recognize the various emotions that a particular tweet expresses semantically.

Keywords: Emotion recognition; text mining; machine learning; unison model; twitter