I've built Webpage classification system with Random Forest from WEKA. I don't known why WEKA Randam Forest's accuracy end up 98% although changes in no of trees. I am still novice in this field, I also would like to know how accuray calculate in it.