基于复杂网络视角的科学文献数据分析

 综述了科学文献数据的复杂网络表现形式，介绍了科学家合作网络和科学引文网络的拓扑结构性质及演化模式和演化机制，概述了论文和科学家的相关评价方法。分析表明，基于复杂网络视角对科学文献数据的分析，能解释很多有意义的研究问题及有趣的现象。

Analysis of scientific literature database from a perspective of complex network

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.