Classification of customer tweets using big data analytics

Abstract

Word of mouth has a great impact on commercial planning and decision-making. Social media is considered as one of the greatest media to spread customer’s opinion about product. Twitter in particular serves as a platform to share people opinion with the words. Decision makers nowadays are seeking analysis approaches on customer tweets to classify whether a customer is satisfied or unhappy. But the enormous number of tweets per seconds and the live streaming of twitter require big data processors in order to support decision-making. In this paper, we propose a recommender system that helps decision makers to fetch customer streaming tweets and classifies their opinion within seconds. We aim to achieve that by applying Naïve Bayes algorithm using big data machine learning approach, Apache Hadoop and Mahout tools are used. The result of our finding is a recommender system that can be used to classify any new customer tweets. The accuracy of the model is 99.39% which promises accurate results in identifying negative or positive customer opinion about a product in a tweet.