In the current world social media plays a vital role to deliver different kinds of emotions in the form of text, emojis, etc. It is very essential that the emotions are very well identified to detect the feelings of the persons who posted it. In this series many kinds of feelings are available in various forms of database repositories. Analyzing all such kind of emotions is very tedious task. This research work applies the Transfer Learning techniques which are used to identify and analyze the emotions produced by different persons in the social media like Twitter, Face book, WhatsApp etc. This approach uses various techniques like Transfer Learning, Convolutional Neural Network (CNN), Deep Learning, Support Vector Machine (SVM) etc. Based on the data this work detects the six different types of emotions namely Happiness, Sadness, Fear, Disgust, Anger, Surprise and Neutral. From this Experimental approach the performance of the chosen algorithm is tested and reported. Finally, Transfer Learning Technique is suggested as the best method for the identification of emotions. Thus, enables psychiatric analysis of patients and interrogation sessions with the accused easy and effective.

Transfer Learning, Emotions Detections, Social media data, Support Vector Machine, Gabor filters, Principal Component Analysis, Convolutional Neural Network