A Survey on Text Classification Techniques for Sentiment Polarity Detection

With the increasing growth and availability of resources, there arises a difficulty in gaining relevant information. Text classification is a mining method to classify each document into a fixed number of predefined classes in order to reduce the length of the text without losing significant information. Opinion mining identifies and extracts subjective information from various sources using techniques such as Natural Language Processing, text analysis and computational linguistics. Opinions provided by individuals and organizations can be utilized for improving market trends and decision making. This paper discusses various text classification techniques for opinion mining such as Bayesian classification, Latent Dirichlet Allocation (LDA) classification, Dynamic Ontology Classification, Novel Algorithm and Genetic Algorithm.