Analysis of human sentiments is a trending topic in this era. The motive of this survey is to determine the emotional state of people or a person by using data mining concepts. Here, the process analyses the state based on keywords provided by the user. Feeling, opinion, emotion or attitude are the Sentiments. By computational identification of a text and processing it, Sentiment analysis can be carried out to find whether the writer's feelings in a specific context is positive, negative, or

more »

... tral. So implementing NLP and to categorize, classify or statistically view the report of analysis from a group of text is what is considered as "sentiment analysis". The motive of this paper is to perform a thorough survey on sentiment analysis algorithms that use machine learning concepts. The topics henceforth explored include various algorithms, quality measurement metrics, feature selection and training methods. Our paper also focuses on the three kinds of machine learning algorithms for Sentiment Analysis-Supervised, Unsupervised and Semi-Supervised Machine Learning Algorithms. I.