Abstract—Users are continuously uploading and sharing photos these days via many photo-sharing services, for example, Flickr and Instagram. The metadata of the photo data usually contains time-, location- and context-related information. This opens a door for researchers to study human social and/or physical behaviors with different perspectives on different data that they may be interested in. The paper first introduces the newly released “100M Yahoo Flickr Creative Commons Images” dataset for research and briefly describes the information contained in the dataset and its potential applications. The objective of this study is to find most visited place in US by Americans based on geo-tagging information retrieved from the dataset. It proposes a method to detect people’s travel patterns as outliers to users’ baseline locations. Detected travelling activities are then attributed to the corresponding geo-grids to build Flickr Tourist Index, from which a ranking of most visited places in US is constructed on a yearly basis. Intuitive map visualizations of Flickr Tourist Index are presented on a US map. The paper also studies trends of ranking changes over years and compares its ranking results with other sources.
Index Terms—Big data, photo-sharing service, pig latin, Flickr tourist index.
Tao Mao is with the School of Information at the University of California, Berkeley, USA (e-mail: tao.mao@berkeley.edu).
[PDF]
Cite: Tao Mao, "Mining One Hundred Million Creative Commons Flickr Images Dataset to Flickr Tourist Index," International Journal of Future Computer and Communication vol. 4, no. 2, pp. 104-107, 2015.
