Exploring Tourist Behavior from Social Media Using Geotagged Photographs

In this paper, we propose an approach for exploring tourists’ behavior based on the extracted dataset from geotagged photographs. We transform the dataset to a suitable format type and apply k-mean clustering to cluster the different tourists’ behavior. After that, the tourists’ behavior of each cluster, especially behavior regards interesting attraction, has been analyzed by statistical test. Entropy also has been employed to measure the diversity of place in each attraction category for presenting each clusters’ characteristics. The results have been demonstrated that k-mean clustering presents the tourists’ behavior into eight groups and each of the tourists prefers conspicuously different attractions.