[4Rin1] Interactive 2

[4Rin1-61] Extracting spots for each tourism resource type using images and location information of Twitter and Flickr in Siem Reap

〇Itaru Morita¹, Kouichi Araki² (1.Nippon Koei Co.,Ltd., 2.Godai Kaihatsu Co.,Ltd)

Keywords:Sightseeing spot extraction, Sightseeing resource classification, Image classification, Twitter, Flickr

In this paper, we introduce a method for extracting secret tourism spots using images and location information obtain from Twitter and Flickr. Our method classifies the images into tourism resources with ResNet and then groups the images in each tourism resource based on location information attached the images with DBSCAN. Finally, in order to extract the secret tourism spots, it evaluates these popularity based on the number of posts and the number of views. We demonstrate that, in Siem Reap, our method extracted several secret tourrism spots, such as natural resource, other than Angkor archeological sites.