that is possible. You could try to predict the correct emotion from the EEG data. Therefore, you will have to understand machine learning classifiers such as logistic regression, Support Vector Machines or Artificial Neural Networks. Using these methods, you could feed in your data and train the models to predict the emotion. These models could then in the future be used to classify emotions in real-time. This could also be built into a GUI.
I would recommend starting of with a thorough investigation of the features to see how well they characterize the emotional states using traditional statistical approaches. After that you can familiarize yourself with machine learning methods. I recommend Andy Ngs Machine Learning Course on Coursera.