Enhancing Hospitality Sentiment Reviews Analysis Performance using SVM N-Grams Method

Abstract

Sentiment analysis or opinion mining is an analysis conducted to derive meaningful information or sentiments contained in an opinion. The use of sentiment analysis has spread in various fields, also exists in the tourism sector. Many tourists are actively reading and writing reviews on travel websites or travel platforms. Whereas in the review information contained useful information for the company or hotel manager, considering that the hospitality industry is very competitive. This analysis produces knowledge about sentiment from the review text data using approaches of n-grams to increase the level of accuracy according to the literature proven. This research uses SVM as a review classification method with positive and negative sentiment. The results of this research indicate an average level of accuracy of 94% which is greater than the level of accuracy in previous research using the same data. In addition, this research shows that the use of SVM as a classification model produces a higher level of accuracy than the Recursive Neural Tensor Network (RNTN).