Sentiment Analysis of Reviews of Tourist Attractions in the Lake Toba Area Using the Naïve Bayes Method

Abstract

Lake Toba is one of the tourism destinations in Indonesia which is the main destination for domestic and foreign tourists. However, the natural beauty of Lake Toba is not enough to develop quality tourist destinations, so analysis needs to be carried out in order to develop tourist destinations that suit tourist needs with the aim of improving the economy from tourism, especially at Lake Toba. One aspect that must be analyzed is comments from tourists who have visited Lake Toba via various platforms. This is very influential for potential future tourists to have a reference for Lake Toba tourism. The analysis process can be carried out by analyzing comments using the Naive Bayes method so that managers of the Lake Toba tourist destination can improve tourist attractions and provide tourist satisfaction and develop various tourism innovations to meet various tourist needs in a sustainable manner. The results of the sentiment analysis of Lake Toba tourist reviews using Naive Bayes detected 31 positive labels, 378 neutral labels and 7 negative labels with an accuracy result of 77.49% from 1260 data, where training data was 1008 and test data was 252 data.