Fig 1.
Bloom’s taxonomy cognitive domain levels.
Fig 2.
Proposed model.
Table 1.
The number of questions in each dataset.
Table 2.
Sample of questions from each dataset.
Table 3.
Weighted F1-measure of different weight cases with collected dataset using KNN, LR and SVM.
Table 4.
Weighted F1-measure of different weight cases with Yahya et al. (2012) dataset using KNN, LR and SVM.
Table 5.
Example of weighting method using classical TF-IDF and modified TFPOS-IDF.
Fig 3.
Example of converting question into a word vector.
Fig 4.
Example of combining word2vec with TFPOS-IDF.
Table 6.
Results of using KNN with TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for the collected dataset.
Table 7.
Results of using KNN with TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for the Yahya et al. (2012) dataset.
Table 8.
Results of LR with TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for the collected dataset.
Table 9.
Results of LR TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for Yahya et al. (2012) dataset.
Table 10.
Results of SVM with TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for the collected dataset.
Table 11.
Results of SVM TF-IDF, TFPOS-IDF, W2V-TFPOSIDF for Yahya et al. (2012) dataset.
Table 12.
Alpha values of t-test.