Twitter based Sentiment Analysis using Hadoop

Abstract

In this paper we propse a joint classification for
tweets using big data and and social networking site twitter.
Today classification using tweets is generally done by splitting
a tweet in to words and not the whole tweet is taken in to
consideration so we thought of introducing a novel approach
where tweets will be classified as whole and not in words. To
enhance the concept we thought of using big data technology such
as Hadoop to help in classification as the tweets that are retrieved
are in large numbers and not easy for a single machine to handle
them. The analyzed information results that are returned will
be assembled together on a single machine and the prediction
returned are in the form of sentiment analysis as each tweet as
a whole will have score and various sentiments will be attached
with them.