A Study on Customer Sentiment Analysis of Commuter Airlines using Twitter Data Mining

Abstract

Sentiment mining has mainly been correlated with analysis of text to establish whether an entity is of positive or, negative polarity. Recently, sentiment mining has been broadened to focus on objects such as distinguishing objective from subjective intentions, and determining the cradles and topics of different opinions expressed in textual formats such as tweets, web blogs, message board reviews, and news. Enterprises can leverage the opinion polarity and sentiment topic recognition to achieve a deeper perspective of the drivers and the overall scope of sentiments. These insights can improve customer service, establish better brand image, and enhance competitiveness. This paper solely examines the trend of analysing consumer feedbacks given in Twitter for various US based Airlines. Researchers are applying data mining and machine learning tools to accommodate different business centric evaluations such as Customer Feedback Assessment, Airline Quality Control, and Consumer Loyalty Prediction etc. Various knowledge building tools (lexicon creation, feature extraction, polarity formation) are emphasized in this study.These are applied to process the raw twitter data into characterized sentiment blocks which determine consumer intuition for choosing desired airlines. Also, external metrics such as weather condition, airline punctuality and service, staff oversight are taken into consideration that impact customer conclusions. This paper succinctly looks through sentiment recognition algorithms with their data processing paradigm and finally possible future directions for improvement are discussed.