<![CDATA[Perum Perhutani is a BUMN (Badan Usaha Milik Negara) whose business portfolio is managing and developing 768 ecotourism in Java and Madura. Perhutani has not been able to develop the tourism objects optimally due to the lack of data and information regarding tourist preferences about tourism objects despite the effort they have done. Sentiment analysis using Multinomial Naive Bayes was carried out to determine tourist preferences from three aspects of tourism objects managed by Perhutani. The three aspects are attractions, accessibilities, and amenities. Reviews obtained from TripAdvisor will be classified into three classes, namely positive, negative, and neutral. The purposes of this study are first, to find out the implementation results of Multinomial Naive Bayes in sentiment analysis process, second, to find out the values ​​of precision, recall, and f1, and lastly to know the results of usability testing from the dashboard. The results of the classification process for the attractions aspect is 69.89% of accuracy. As for accessibilities and amenities, the accuracy values ​​are 61.67% and 63.01% for each aspect. Information about of tourist sentiment is displayed on a dashboard which is then tested by Perhutani using SUS (System Usability Scale) and gets a score of 65. That result means the dashboard has “Good†rating and in range Marginal (High) for acceptability.]]>
Copyrights © 2021
