Both sellers and buyers heavily depend on the
opinions of customers in purchasing and selling products online.
When it comes to text-based data, sentiment analysis of user
reviews has become a prominent facet of machine learning. Text
data is generally unstructured which makes opinion mining very
challenging. A wide array of pre-processing and post-processing
techniques need to be applied. But the major challenge is
selecting the right classifier for the job. Naïve Bayes algorithm is
a commonly used machine learning classifier when it comes to
opinion mining and sentiment analysis. The focus of this survey
is to observe and analyze the performance of Naïve Bayes
algorithm in sentiment analysis of user reviews online. Recent
research from a wide array of use-cases such as sentiment
analysis of movie reviews, product reviews, book reviews, blog
posts, microblogs and other sources of data have been taken into
account. The results show that Naïve Bayes algorithm performs
exceptionally well with accuracies between 75% to 99% across
the board.