Machine Learning Approach for Sentiment Analysis

Machine learning algorithms have been widely used for sentiment analysis [66]. The bag-of-words (BoW) representation is commonly used for sentiment analysis [63, 93]. BoW method assumes the independence of words and ignores the importance of semantic and subjective information in the text. All the words in the text are considered equally important. The BoW representation is commonly used for sentiment analysis, resulting into high dimensionality of the feature space. Machine learning algorithms reduce this high-dimensional feature space with the help of feature selection techniques which selects only important features by eliminating the noisy and irrelevant features. Recently, machine learning-based sentiment analysis models are gaining prominence in the field [66].