Enhancement of co-authorship networks with content-similarity information

Co-authorship networks contain a wealth of information from the collaboration of the researchers, the research areas of individual authors to the detection of research communities and much more. Its applications in the field of learning and research include expert detection in a research area to the identification of institutions that involve in research on any particular area. With the exponential rise in research in the recent past and the enormity of the Co-authorship networks, research on mining knowledge from them has been an active area of research. We propose an enhancement to the co-authorship network. New edges are introduced based on the similarity within abstracts. This model would help in identification of potential research synergies even when the authors don't share a common edge in the original co-authorship graph. The initial experiments show that the modifications in the graph improve the co-authorship network in a semantic perspective.