LSTM and Bi-LSTM Deep Learning Technique for better Tourism Services in future by analyzing Hotel Reviews

Main Article Content

A. Ms. Harsh Arora, B. Dr. Mamta Bansal

Abstract

One of the worthwhile aspects of the arena economy is tourism. Tourism plays a very important part in the economy of any country. Tourists make extensive use of mobile devices and e-tourism services throughout their journeys. On the internet, there is a significant growth of opinions and reviews, which always contain product and service evaluations, as well as thoughts about events and people. Online booking of hotels has been tremendously increased in past few years using e-commerce technologies. Tourist’s reviews are having great importance to know the positive and negative points of current tourism industry so that we can work on these gaps to improve the current tourism services. It is possible to use the results of tourist’s behavior analysis to improve the existing tourism services. In this paper, in order to detect customer sentiments in form of positive and negative reviews, the sentiment analysis has been used. In context of that, the online data in form of hotel reviews is fed to classical algorithms or techniques of machine learning and same data is given as input to deep learning method called LSTM and further the results are compared between these two where LSTM outperformed traditional machine learning algorithms like the Nave Bayes and Random forest algorithms. Here to improve the text classification, polarity analysis has been done by taking data of hotel reviews given by customers. Using the Long Short-Term Memory (LSTM) deep learning method, the sentiment expression is employed to categorize the polarity of the textual content evaluation on a scale of dreadful to favorable. The purpose of this research is to depict how accurate and efficient results come in to picture when we use deep learning method rather than traditional machine learning methods. While using sentiment analysis, deep learning LSTM proposed model delivers higher accuracy in results of hotel reviews.

Article Details

Section
Articles