Analysis of Opinion Mining on Social Media Data Streams using Hadoop

Abstract

Twitter is a social networking site in which the data to be processed is in rich amounts and which can be structured, semi-structured and unstructured data streams. Opinion mining over the Twitter offers organizations a fast and effective way to monitor the feelings of public towards their services. It focuses on predicting the polarity of words and then classifies them into positive and negative feelings with the aim of identifying attitude and opinions that are expressed in any form or language. Bian et al. (2014) explains how the Map -Reduce paradigm can be applied to existing Naïve Bayes algorithm to handle a large number of tweets. All these approaches say about the real-world data sets at its accuracy level by using Hadoop File System. This paper analyses all the above methods comparatively.