Scientific literature data cover the complete information of papers and authors. Facing the massive scientific literature data, traditional statistical analysis methods cannot fully explore the information hidden behind the data without the help of other analysis methods. The interactions in scientific literature data, such as citation between papers and co-authorship between scientists, allow for the construction of different forms of complex networks (citation networks, collaboration networks, etc.), which can allow us to distinguish the effective information hidden in the scientific literature data based on network analysis. This paper summarizes the complex network forms of scientific literature data and highlights the topological properties, evolution patterns as well as evolution mechanisms of scientific collaboration networks and scientific citation networks. As impact evaluation of papers and scientists has attracted so much attention from researchers for a long time,, we also briefly summarize the related evaluation methods of papers and scientists. From the perspective of complex network, it can also explain many meaningful questions and interesting phenomena in scientific literature data, such as the shift of scientists' research interests and the sleeping beauties. In the future, the method of network analysis must be able to achieve more abundant research results in mining scientific literature data.

scientific literature data  /  complex network  /  evolution  /  evaluation