Opinion mining from online reviews in Bali tourist area

Abstract

Bali Island is the most popular tourist destination in Indonesia. Bali needs to make continuous quality improvements of its tourism industry by devoting particular attention to the hotel as an integral part of tourism. Through hotel user reviews, hotel managers gained insight about the hotel condition that was perceived by the users. based on online reviews in Tripadvisor.com, this study used text mining approach and aspect-based sentiment analysis to obtain hotel user opinion in the form of sentiment. Aspect-based sentiment analysis is able to provide information that is not provided by the typical sentiment analysis. To perform these tasks, this study tries to apply the Recursive Neural Tensor Network (RNTN) algorithm, which was commonly used for classifying sentiment in sentence level. With the average accuracy of 85%, the proposed algorithm performed well in classifying the sentiment of words or aspects. Moreover, the output can be used for evaluation in improving the quality of the hospitality industry as well as supporting the tourism industry in Indonesia.