Bài báo khoa học

In scientific research, researchers collaborate on research and publication of scientific papers. The connection of scientists in scientific papers is considered a special social network and is called a co-authorship network. The collaborative recommendation problem in the co-authorship network is an exciting and meaningful problem to help expand the research network and promote the development of scientific research. In this article, we approach the co-authorship recommendation problem according to the classification problem through the co-authorship candidate table. Accordingly, this article proposes a semantically rich fuzzy membership function for the FSVM-CIL algorithm to solve the co-authorship recommendation problem. Experimental results on co-authorship datasets with different characteristics show that our algorithm is more efficient than SVM, WSVM, and FSVM-CIL classification algorithms.