Customer Review Analytics for Business Intelligence

Abstract

With the advent in Big Data Analytics and Machine Learning newer opportunities are created for Business Organizations to analyze customer reviews for Business Intelligence. Nowadays customers share their opinions online through social media platforms and review sites. This paper proposes a Novel Machine Learning based Ensemble Classifier to classify review sentences as praise or complaint and Text Analytics-based approach is developed to gain valuable insights from praise and complaint sentences. The proposed Ensemble Classifier gives 79.83% accuracy with bigram and hybrid features as compared to baseline classifiers such as Multinomial NB, Decision Tree, etc. The Praise and Complaint sentences are further analyzed and top-ranked praises/ complaints are identified, their word clouds are displayed from Business Intelligence point of view. The proposed approach can be used to analyze the vast amount of customer reviews to understand customer's satisfaction or concern's based on praises and complaints expressed in unstructured reviews.