This event has ended. View the official site or create your own event → Check it out
This event has ended. Create your own
View analytic
Saturday, September 14 • 4:46pm - 6:31pm
“Analyzing spatial, social, and semantic dimensions of user interactions with collections on Flickr”

Sign up or log in to save this to your schedule and see who's attending!

In 2008, the Library of Congress (LC) posted a collection of historical photographs on Flickr (http://www.flickr.com/photos/library_of_congress/), a photo sharing and social tagging service, with an intention of engaging the Web 2.0 community. Our project investigates a sample of 4500 records from the LC Flickr project with the purpose of understanding how Web 2.0 communities interact with the collections on Flickr. The goal is to develop a holistic picture of these interactions and analyse them from multiple perspectives: spatial, social, and semantic. The initial exploratory findings reveal that collections of photographs are enhanced by communities of taggers and commenters. These communities contribute different kinds of knowledge to image collections. Taggers add tags; commenters contribute local knowledge, opinions, and sentiments. In anthropology, local knowledge is defined as knowledge strongly rooted in particular places and reflects personal and emotional awareness of an area (Geertz, 1983).

In this project we are closely investigating these communities with the purpose of determining who plays the key role in knowledge creation, and what collections benefit from such knowledge enhancements the most. To achieve this objective, we are planning to examine collections, users, and their contributions by means of social network, geographic, and semantic analysis. The application of SNA in this scenario is somewhat unusual. Commonly, SNA is used for analysis of communities of people. Recent studies, however, have used SNA for analysis of other objects, namely: diseases (Goh et al., 2007), recipes (Teng, Lin, & Adamic, 2012), and images (Cha, Mislove, & Gummadi, 2009). Therefore, we assume that application of SNA to user interactions with images is justified. For SNA we mapped image-tagger and image-commenter relationships. To differentiate comments, we semantically analyzed and then categorized comments by content: local knowledge (that includes stories, opinions, questions, and corrections), and sentiments. This allowed us to see the differences in the structure of interactions in different communities. Our initial findings demonstrate that user comments include more local knowledge statements than sentiments; there are more local knowledge commenters than taggers, but fewer than those who write sentiments. User nodes in the local knowledge community are larger than in the sentiment community. By size, nodes differ significantly: in the local knowledge community, the most prolific user has 330 relationships with images, but the most prolific user in the sentiment community has only 70 relationships. The overlap between local knowledge and sentiment communities is not large: those who provide local knowledge hardly provide any sentiments.

With geographic analysis, we intend to infer users’ expertise in local knowledge of places and people shown in images. To carry out such analysis, we will map locations featured in images and locations of contributors. The distances between images and contributors may provide clues to the degree of users’ expertise in the knowledge of a place or an event featured on a photograph.


Cha, M., Mislove, A., & Gummadi, K. P. (2009). A measurement-driven analysis of information propagation in the flickr social network. In Proceedings of the 18th international conference on World wide web (pp. 721-730). ACM.

Geertz, C. (1983). Local knowledge: Further essays in interpretive anthropology. New York: Basic Books.

Goh, K. I., Cusick, M. E., Valle, D., Childs, B., Vidal, M., & Barabasi, A. L. (2007). The human disease network. Proceedings of the National Academy of Sciences, 104(21), 8685-8690.

Teng, C. Y., Lin, Y. R., & Adamic, L. A. (2012, June). Recipe recommendation using ingredient networks. In Proceedings of the 3rd Annual ACM Web Science Conference (pp. 298-307). ACM.

avatar for Jihee Beak

Jihee Beak

University of Wisconsin-Milwaukee, United States

Olga Buchel

University of Western Ontario, Canada

Inkyung Choi

University of Wisconsin-Milwaukee, United States

Margaret Kipp

University of Wisconsin-Milwaukee, United States

Diane Rasmussen

University of Western Ontario, Canada

Saturday September 14, 2013 4:46pm - 6:31pm
Rowe Atrium

Attendees (4)