Prediction of favourite photos using social, visual, and textual signals

摘要: This paper focuses on the prediction of users' favourite photos in Flickr. We propose a multi-modal, machine learned approach that combines social, visual and textual signals into single system. Although each individual user has different motivations for calling photo favourite, we show textual, visual, social modalities effectively capture needs most active Flickr users. use gradient-boosted decision trees (GBDT) with mod least squares loss function classification user's photos, evaluate performance our classifier respect to various combinations thereof. By using combination GBDT creates highly effective classifier. The addition features allows us significantly increase recall, slight trade off precision.