With the success of Web 2.0 applications, various social media websites have been established and become tremendous assets for supporting critical business intelligence applications. The knowledge gained from social media websites can not only meet the objectives of businesses offering them but also help the development of novel and effective services that are better tailored to users’ needs. In this study, we concentrate on proposing effective personalized recommendation systems in social bookmarking websites. The proposed technique exploits the valuable and unique resources, i.e., folksonomy, in social bookmarking websites to design a profile expansion mechanism which enhances the effectiveness of personalized recommendation. According to our empirical evaluation results using the data of a leading scientific reference sharing website, i.e., CiteULike, our proposed technique significantly outperforms its benchmark techniques.