A study on Sentiment Analysis Approaches

Text mining is attracting considerable interest due to the enhancement of resources on the Internet. Sentiment Analysis is recognized as being the most important of natural language processing. It is a process automatic extraction of knowledge which exploits the unstructured data automatic from the reviews about specific topic. Several points of view have been put forward in the literature. The aim of Sentiment Analysis is extraction of opinions from the web sites and categorizing the polarity of review in terms of positive, negative. To achieve this aim, we illustrate the recent trend of researches in the Sentiment Analysis and its related domains. In this paper, several techniques of Sentiment Analysis on various datasets were considered. In the end, we illustrate an analogically evaluation of such techniques in terms of accuracy.