Tag-based personalized image ranking in event browsing

Many image sharing websites, e.g. Flickr, Google+, allow users to upload images as an event, and users can browse the images others uploaded as events. The fact that people usually browse only the first few images of an event then decide whether the event is what they want makes us believe that it is necessary to present those images people favor on the very first position for each event. Here we propose a new tagbased personalized image-ranking algorithm in event browsing such that it gives image higher score if it: a) is important in the event, b) matches user’s preference. c) matches user’s query. To this end, we first adopt a local matching model to assign images an original score based on whether this image satisfies user’s query and preference. We then propose a global ranking model to take the local scores as initial values and make the salience scores iteratively smooth with respect to all images returned from the events of the query.