In our thesis we have worked to analyses text short answers then predict the score accordingly by using different extracted features. In our research we have used around 1700 data for each dataset and are scored by two ...