An Ensemble Sentiment Classification System of Twitter Data for Airline Services Analysis

Abstract

In airline service industry, it is difficult to collect data about customers' feedback by questionnaires, but Twitter provides a sound data source for them to do customer sentiment analysis. However, little research has been done in the domain of Twitter sentiment classification about airline services. In this paper, an ensemble sentiment classification strategy was applied based on Majority Vote principle of multiple classification methods, including Naive Bayes, SVM, Bayesian Network, C4.5 Decision Tree and Random Forest algorithms. In our experiments, six individual classification approaches, and the proposed ensemble approach were all trained and tested using the same dataset of 12864 tweets, in which 10 fold evaluation is used to validate the classifiers. The results show that the proposed ensemble approach outperforms these individual classifiers in this airline service Twitter dataset. Based on our observations, the ensemble approach could improve the overall accuracy in twitter sentiment classification for other services as well.