A Novel Approach on Social Re-Ranking by Tag Based Image Search

Abstract

Social media sharing websites like Flickr allow users to define images with free tags, which significantly contribute to the development of the web image retrieval and organization. Tag based image search is an important technique to find images contributed by social users in such social websites. However, how to make the top ranked result suitable and with diversity is challenging. In this paper, we propose a social re-ranking system for tagbased image retrieval with the consideration of image’s relevance and variance. We aim at re-ranking images according to their visual information, semantic information and social hints. The primary results include images contributed by different social users. Usually each user contributes several images. First we sort these images by inter-user re-ranking. Users that have higher contribution to the given query rank higher. Then we sequentially implement intra-user reranking on the ranked user’s image set, and only the most related image from each user’s image set is selected. These selected images comprise the final retrieved results. We build an inverted index structure for the social image dataset to accelerate the searching procedure.