Abstract | ||
---|---|---|
This study employed text data mining to demonstrate the reliability of identifying tourist needs from travel reviews by comparing the results of a traditional tourism survey with the attitudes expressed in travel reviews. In this study, we focused our analysis on tourist satisfaction and adopt the results of a governmental satisfaction survey implemented in Hokkaido, Japan (n = 1709) for referential statistics. We used manual techniques to extract attitudes from 1058 samples of reviews (in English, Simplified Chinese, and Traditional Chinese) posted on TripAdvisor by tourists from seven different regions. By calculating the Pearson’s r, we found a (strong) positive correlation between attitudes in reviews and the satisfaction rates recorded in the guest survey in six out of seven regions (p < 0.05). Meanwhile, Fisher’s exact tests showed that the percentages of positive reviews are different from the satisfaction rates in the guest survey. On the other hand, the percentages of combined positive and neutral reviews are numerically similar to the satisfaction rates registered in the guest survey. Further validation could be considered, along with a comparison of other statistics for tourist satisfaction, additional review samples, and the help of a developed automated analysis method. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1007/s40558-019-00144-3 | Information Technology & Tourism |
Keywords | Field | DocType |
Text mining, Needs investigation, Cross language, Pearson correlation coefficient | Pearson product-moment correlation coefficient,Computer science,Tourism,Positive correlation,Statistics,Marketing | Journal |
Volume | Issue | ISSN |
21 | 3 | 1098-3058 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Shuang Song | 1 | 17 | 13.19 |
Hidenori Kawamura | 2 | 42 | 18.22 |
Junichi Uchida | 3 | 0 | 0.34 |
Hajime Saito | 4 | 0 | 2.37 |