Understanding human emotion has been an important topic of interest for researchers for a long time. There exist many passive forms of emotions that are not visible clearly by the gestural feed but reflect the changes in the body’s internal parameters e.g., temperature, sweat, blood pressure, pulse rate, etc. However, passive emotion has been tried to identify with the help of ECG and EEG data. Though those techniques have many limitations, very few works address the issue with the help of Galvanic Skin Response (GSR). In this work, we specifically address the identification of the passive nature of emotion. To do it, we have used GSR data to identify the seven emotions (Anger, Fear, Sadness, Disgust, Happiness, Surprise, and Mixed) precisely where each emotion has been measured on a scale of five points (very low, low, neutral, high, very high). We use the YAAD dataset for our experiments. We use 10 machine-learning models to do this multi-class emotion identification. It shows more than \(70\%\) accuracy which outperforms the existing methods which deal with the GSR data.