Abstract: In this paper, we studied the performances of models when features that are not very common are not included to train the models used for automated evaluation of descriptive answers which ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback