Author Name Disambiguation Using Two Stage Clustering

Differentiating different authors with same or similar names plays an important role in various digital library and other related services. For the effective use of these author services it is important that the authors be profiled in an efficient manner. Without a proper disambiguation method in place it is very difficult to profile different authors. The quality of the services provided by the digital library or any other similar application depends directly on the quality of disambiguation mechanism in place. Soft Computing paradigm offers variety of techniques that can be employed to improve such an important task. These techniques can be used to efficiently cluster different authors and their related publications. In this paper we propose a two stage clustering framework for disambiguating author names. The clustering results of the first stage are used to further improve the quality of clustering in the second stage. We explore and investigate the use of a mix of soft and hard clustering techniques for profiling and disambiguation of authors and in turn their publications.