การวิเคราะห์ผลสะท้อนกลับด้านการท่องเที่ยวจากบทวิจารณ์ออนไลน์ กรณีศึกษาจังหวัดภูเก็ต

Abstract

“Phuket”, a small province in Thailand, is the main source of income for tourism in Thailand because Phuket attractions include natural attractions, recreational activities and a variety of food that can attract both domestic and internation tourists enormously. The world today The Internet plays a huge part in finding information and planning a trip. For example, travel related websites and travel media use a presentation style that contains a section of online reviews as the main for exchanging idea, opinion or sharing the experiences that tourists have actually experienced. Accessing such information becomes a great tool for travelers while making their decisions. However an issue with online reviews includes the amount of information, the language used by the reviewers and the details that the reviewer want to express. Therefore, this research aims to study the behavior of tourists when traveling around attractions in Phuket. All data collected from the Tripadvisor website between 2010-2020 were used. There were 190 Phuket attractions with the total of 76,183 reviews. Techniques include the Association Rule using the FP-Growth algorithm, Text mining using text frequency analysis and Part of Speech – POS for finding the thing that tourists need or frequently mentioned, Term Frequency – Inverse Document Frequency (TF-IDF) techniques for considering the importance of reviews, and Sentiment Analysis for finding the real feeling of tourists, The final outcome will be summarized into a Dashboard. Tourists, related officials or entrepreneurs in Phuket can use the dashboard in order to be prepared for each group of tourists at different locations and at different time periods as efficiently as possible.