Personalizing image tags is a relatively new and growing area of research,and in order to advance this research community, we must review and challengethe de-facto standard of defining tag importance. We believe that for greaterprogress to be made, we must go beyond tags that merely describe objects thatare visually represented in the image, towards more user-centric and subjectivenotions such as emotion, sentiment, and preferences. We focus on the notion of user preferences and show that the order that userslist tags on images is correlated to the order of preference over the tags thatthey provided for the image. While this observation is not completelysurprising, to our knowledge, we are the first to explore this aspect of usertagging behavior systematically and report empirical results to support thisobservation. We argue that this observation can be exploited to help advancethe image tagging (and related) communities. Our contributions include: 1.) conducting a user study demonstrating thisobservation, 2.) collecting a dataset with user tag preferences explicitlycollected.
展开▼