Sentiment Analysis on Tweets for Trains Using Machine Learning

Abstract

Sentiment analysis is a popular theme in the natural language processing (NLP) domain. People at present share their stay experience in restaurants, shopping malls, hotels and their travel experience in taxis, buses, trains and airplanes. Online social media provide a platform for the people to share their experiences of stay and travel in the form of text, images and videos. Twitter is one of the popular and well known social media platforms across the world. In this study, we are using tweets data in respect to comfort services in Indian long route superfast trains. This tweet data is used to analyze the hidden sentiments using machine learning techniques such as support vector machines (SVM), Random forest (RF) and back propagation neural networks (BPNN). The results show that BPNN provides high accuracy with more training on the data. The results achieved from SVM and RF was also satisfactory but BPANN won the race with more training on the data.