We propose to use real-time single-channel EEG signal to quantify user enjoyment level elicited by media content. Selected time-frequency components from single-channel frontal EEG were extracted and formed a statistical multivariate model predicting user enjoyment level. Frequency components from Theta, Alpha and Beta bands at different time moments were selected. We found robust model performance during 10-folds cross-validation with 100 repetitions. A high correlation of around 0.8 between predicted and actual enjoyment level of subjects was achieved. Considering various factors of the selected features, we found an important role of alpha as the emotional component, and beta as the cognitive component involved in the complex enjoyment processes. Also, in accordance with the peak-end rule,
feature from latter part of the video seems to create a large influence to the overall experience. From all of these results, we implement real-time EEG-based detection system for media user enjoyment with single EEG channel.