Table 6. Performance result of classification algorithms (F1-measure).

Model	Original dataset (%)	Converted dataset (%)
CNN	79.7	80.8
RNN	55.9	70.5
CNN + Doc2vec	93.7	88.3
kNN	80.3	80.1
Logistic regression	89.7	86.9
Naïve Bayes	88.1	87.2
N-gram language model	87.8	85.8