Lack of Training data in Sentiment Classification: Current Solution
Mohammad Sadegh Hajmohammadi, Roliana Binti Ibrahim · 2012
In recent years, sentiment classification has attracted much attention from natural language processing researchers. Most of researchers in this field consider sentiment classification as a supervised classification problem and train a classifier from a large number of labelled documents. . Unfortunately, in some language other than English, a reliable and sufficient labelled data is not always available and manually labelling a reliable and rich training data is very time-consuming. Until now, researchers have developed several techniques to the solution of the problem. This paper try to cover some techniques and approaches that be used in this area.