Sentiment Analysis of Twitter Data In Hadoop Using Naive Bayes And Fuzzy C Means Clustering

Social networking is one of the main platforms responsible for the increased amounts of data generation from the users. People across different geographical regions share their thoughts and put their opinions on the micro-blogging sites. One of the most popular micro blogging sites is Twitter where people share their reviews in the form of tweets. Due to concise and short limits of tweets, it is easier to analyze and thus extract valuable outcomes from tweets. The tweets also provide varied content of sentiments and opinions about the current technologies and affairs. Sentiment analysis is the process of analyzing various opinions and reviews given by people. Sentiment Analysis is the process which tends to understand these opinions and categorize them into positive, negative and neutral categories.

In this paper, the authors propose a concept for sentiment analysis that will help to classify various tweets on the basis of sentiment polarity. The streaming dataset from twitter will be stored in HDFS clusters which will be mined later using Naïve Bayes and Fuzzy C Means Algorithms to improve scalability and accuracy of various performance metrics and thus help any organization to formulate various strategies to promote their work process.