TY - JOUR
T1 - Analyzing Google trends with travel keyword rankings to predict tourists into a group
AU - Chang, Jui-Hung
AU - Tseng, Chien Yuan
PY - 2019/1/1
Y1 - 2019/1/1
N2 - This study explored the correlation between tourism-related popular words in search engines and the number of tourists, which is a topic worth discussing for the tourism industry. When individuals decide to sign up for a tour planned by travel agencies, they often have no idea whether the minimum number of participants will be reached until it is confirmed a few days before the departure, which is of great inconvenience when arranging itineraries. Hence, predicting whether the minimum number of participants in a tour can be reached is closely related to both the tourism industry and the tourists. In this regard, the number of Taiwanese traveling to Japan was predicted based on the popularity of keywords concerning travel in Japan searched on Google Trends. The scores of popular words concerning travel in Japan on the Google search engine and words concerning travel in Japan mentioned in tourism articles in e-news networks were also summarized. The experimental results indicated that the popularity of tourism keywords on Google was highly correlated to the number of Taiwanese tourists traveling to Japan. After the number of Taiwanese traveling to Japan within the following month was classified by the ANN model, the mean square error reached 0.13. Furthermore, by using the data of travel agencies in Taiwan to match the Google Trends data, the research predicted whether tours to Hokkaido would reach the minimum requirement for participants. The prediction accuracy of the ANN model was 68%.
AB - This study explored the correlation between tourism-related popular words in search engines and the number of tourists, which is a topic worth discussing for the tourism industry. When individuals decide to sign up for a tour planned by travel agencies, they often have no idea whether the minimum number of participants will be reached until it is confirmed a few days before the departure, which is of great inconvenience when arranging itineraries. Hence, predicting whether the minimum number of participants in a tour can be reached is closely related to both the tourism industry and the tourists. In this regard, the number of Taiwanese traveling to Japan was predicted based on the popularity of keywords concerning travel in Japan searched on Google Trends. The scores of popular words concerning travel in Japan on the Google search engine and words concerning travel in Japan mentioned in tourism articles in e-news networks were also summarized. The experimental results indicated that the popularity of tourism keywords on Google was highly correlated to the number of Taiwanese tourists traveling to Japan. After the number of Taiwanese traveling to Japan within the following month was classified by the ANN model, the mean square error reached 0.13. Furthermore, by using the data of travel agencies in Taiwan to match the Google Trends data, the research predicted whether tours to Hokkaido would reach the minimum requirement for participants. The prediction accuracy of the ANN model was 68%.
UR - http://www.scopus.com/inward/record.url?scp=85071185516&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85071185516&partnerID=8YFLogxK
U2 - 10.3966/160792642019012001023
DO - 10.3966/160792642019012001023
M3 - Article
AN - SCOPUS:85071185516
VL - 20
SP - 247
EP - 256
JO - Journal of Internet Technology
JF - Journal of Internet Technology
SN - 1607-9264
IS - 1
ER -