Title: Current Awareness & Call for abstractors for eLucidate
Abstract:the researchers examined metadata provided by users of the photo-sharing website Flickr.Their goal was to see what they could learn about how users classify content, and if any of that knowledge could...the researchers examined metadata provided by users of the photo-sharing website Flickr.Their goal was to see what they could learn about how users classify content, and if any of that knowledge could potentially be applied to our own systems.The researchers examined 3,000 photos from 879 individual users, 300 photoset (album) descriptions, and discussions from 200 group photo pools.Overall, the team found that Flickr users focused primarily on identifying people, places and activities in their photos.They also found many users who did not use tags at all on their photos, and instead relied on photoset descriptions to provide metadata.For those who did use tags, they were used both individually (to identify a particular friend) and collectively (to identify a public event or place, for example).This data was mapped against a previous photo sorting and identification study, and the authors provide a brief analysis.In addition, they also briefly compared Flickr's group categories and the guidelines for posting within those groups to a handful of current metadata frameworks.Continuation of the research should yield some interesting, more concrete, recommendations.[AC]Read More