Feature Selection for Sentiment Analysis by using SVM

Sentiment analysis depends on feature selection methods to approaches that use general statistical measures where features are selected on empirical evidence. Empirical evidence (research) is a way of gaining knowledge by means of direct and indirect observation or experience. there are new features selection schemes that use a content and syntax model that is used to automatically learn a set of features in a review document and removing the entities that are being reviewed from the subjective expression that describe those entities in terms of polarities.