Recommend Social Network Users Favorite Brands

Abstract

With the development of social network and image sharing websites, users are willing to upload their favorite photos on the websites and assign them some texts to describe the image content. Thus we can capture their interest by these photos and corresponding texts, and recommend relevant brands based on user’s interest. This paper proposes a novel brands recommendation approach for social network users based on their browsing images and labeled texts. Firstly, we enrich the uploaded image’s texts by image annotation approach. Secondly, we build brand tree from the collected datasets. And then, we recommend brands by scalable brand mining based on tree structure. Finally, we conduct a series of experiments on real Flickr users. The experiment results show the effectiveness of our approach.