Predicting Polarity of Tourists Reviews using LSTM Deep Learning Model over Machine Learning Classical Approach to increase the accuracy in Text Classification

Abstract

^{The objective of sentiment analysis is to find positive and negative reviews. Hotel reviews have been analyzed in this research
paper using machine learning algorithms and these are compared with most efficient deep learning method based on LSTM.
The idea is to use the concept of text classification in the form of customer reviews. Although there are many machine learning
techniques available for text classification but here two most important methods of machine learning have been used which
are Naïve Bayes algorithm and another is Random Forest algorithm. These techniques are further compared with LSTM based
deep learning technique. The goal is to develop a deep learning model that uses the LSTM technique to work on hotel reviews
in the context of online tourism and outperforms earlier machine learning techniques. This methodology will aid the tourism
industry in growing its business by studying consumer hotel reviews. Sentiment analysis is widely used in business domains to
improve products and services by understanding customer opinions about these services and this is the case with hotel reviews
analysis using LSTM to obtain efficient and clear results that will undoubtedly help the tourism industry in better understanding
the customer opinions so that in the future, the tourism industry can deliver better services to customers.
Keywords: Machine learning, Deep Learning, Hotel Reviews, Online Tourism, LSTM}