An Image is Worth More than a Thousand Favorites: Surfacing the Hidden Beauty of Flickr Pictures

The dynamics of attention in social media tend to obey power laws. Attention\nconcentrates on a relatively small number of popular items and neglecting the\nvast majority of content produced by the crowd. Although popularity can be an\nindication of the perceived value of an item within its community, previous\nresearch has hinted to the fact that popularity is distinct from intrinsic\nquality. As a result, content with low visibility but high quality lurks in the\ntail of the popularity distribution. This phenomenon can be particularly\nevident in the case of photo-sharing communities, where valuable photographers\nwho are not highly engaged in online social interactions contribute with\nhigh-quality pictures that remain unseen. We propose to use a computer vision\nmethod to surface beautiful pictures from the immense pool of\nnear-zero-popularity items, and we test it on a large dataset of\ncreative-commons photos on Flickr. By gathering a large crowdsourced ground\ntruth of aesthetics scores for Flickr images, we show that our method retrieves\nphotos whose median perceived beauty score is equal to the most popular ones,\nand whose average is lower by only 1.5%.\n