User similarity-based gender-aware travel location recommendation by mining geotagged photos

Abstract

The popularity of camera phones and photo sharing websites, e.g., Flickr and Panoramio, has led to huge volumes of community-contributed geotagged photos, which could be regarded as digital footprints of photo takers. Thus, mining geotagged photos for travel recommendation has become a hot topic. However, most existing work recommends travel locations based on the knowledge mined from photo logs (e.g., time, location), and largely ignores the knowledge implied in the photo contents. In this paper, we propose a geotagged photos mining-based personalised gender-aware travel location recommendation approach, which considers both photo logs and photo contents. Firstly, it uses an entropy-based mobility measure to classify geotagged photos into tour photos or non-tour photos. Secondly, it conducts gender recognition based on face detection from tour photos. Thirdly, it builds the gender-aware profile of travel locations and users. Finally, it recommends personalised travel locations considering both user gender and similarity. Our approach is evaluated on a dataset, which contains geotagged photos taken in eleven cities of China. Experimental results show that our approach has the potential to improve the performance of travel location recommendation.