Sentiment Analysis for Airline Tweets Utilizing Machine Learning Techniques

Abstract

The airline industry is a very competitive market. In the past two decades, the industry has expanded and still expanding its routes domestically and internationally with the web’s improvement and providing better services driven by competition. The Twitter boom has boosted investigation in this area owing to its potential applications, and survey gatherings offer a rich wellspring of feeling information for sentimental data mining. It is necessary to see the data and use mechanism that examination of such information and the territory of sentiment analysis has been developed. The essential assessment is to discover the extremity of the content and order it into positive, negative or neutral sentiment tweets. Therefore, the focal point of this research paper is to perform estimation investigation using machine learning (ML) methods for sentiment analysis. This work endeavoured to use three different ML techniques for the undertaking of estimation examination. The ML tests are performed, while utilizing US airline Twitter informational index is caught from Kaggle. The efficacies of three ML grouping strategies like decision tree, SVM and neural networks are evaluated, and there is consideration about accurate information. The neural network-based approach has perceived a most noteworthy accuracy (75.99%).