Analysis of social information for author recommendation

Researchers in almost all scientific disciplines rely heavily on the collaboration of their colleagues. Throughout his or her career, any researcher will build up a social academic network consisting of people with similar scientific interests. A recommendation system could facilitate the process of identifying and finding the right colleagues, as well as pointing out possible new collaborators. As a researcher's reputation is of great importance, the social information gleaned from citations and reference data can be used to cluster similar researchers. Web services, such as social bookmarking systems, provide new functionalities and a greater variety of social information - if exploited correctly; these could lead to better recommendations. The following describes, by way of example, one approach to author recommendation for social networking in academia.