Sentiment Analysis Using K Means Clustering on Microblogging Data Focused on Only the Important Sentiments

The aim of this paper is to perform sentiment analysis on a microblogging data (Twitter data is taken for the experiment) using k means clustering algorithm. The main intention here is to study the conditions that may favor or not, sentiment analysis using k means clustering considering only two major sentiments as positive and negative. The other sentiments which are not of major interest are made into another cluster and the accuracy of the clusters thus formed is analyzed and meaningful conclusions from the same can be obtained.