A Semantic-Driven Music Recommendation Model for Digital Photo Albums

Jiansong Chao, Haofen Wang, Wenlei Zhou, Weinan Zhang, Yong Yu

Abstract: Digital photo album softwares like iPhoto have enjoyed great popularity for years. These years, online photo album services (e.g., Flickr and Picasa) have been becoming more and more popular with the development of social Web. In this paper, we present a semantic-driven model to recommend music for photo albums automatically. In particular, we exploit semantic data to represent both images and music. Furthermore, we leverage mining techniques to capture semantic relatedness between these different types of multimedia data, which is the essential step for recommendation. In the experiment, our method achieved a performance of about 68% satisfaction measured by participants’ feedback.