Bag of words for feature extraction and then we use a svm for th classification
1 Gather all the words that exist in the dataset and apply the morphological processing.
2 Remove the "Stopwords".
3 Remove the punctuation marks, and make a sorted list where each word appears once.
4 Replace every review with a feature vector.
At test set
Predicted Label | ||
---|---|---|
Negative | Positive | |
Negative | 362 | 91 |
Positive | 102 | 345 |
Accuracy | Presicion | Recall | Fmeasure | Spesificity | |
---|---|---|---|---|---|
SVM | 0.78 | 0.79 | 0.77 | 0.78 | 0.79 |