PhotoSense: emergent semantics based approach to image annotation

Tagging of images using descriptive keywords (tags), contributed by ordinary users, is a powerful way of organizing them. However, due to the richness of the image content, it is often difficult to choose tags that best describe the content of the image to the viewing audience and ensure access to the image. In this paper, we present a novel tagging framework based on the theory of emergent semantics to assist the user in the tag selection process. Our idea is to enrich the current "looking at" experience of tagging with the "looking for" experience of searching. We describe the design of our approach along with a preliminary user study conducted with a prototype Flickr application.