科学知识网络自相似性的实证研究

Empirical Study on the Spatial Self-similarity of Scientific Knowledge Network

科学计量学的研究都是以科学知识的自相似性作为理论假设的,尤其是科学知识图谱更是以科学文献等在空间上的自相似性为前提,因此对科学知识网络自相似性的检验与证明是必不可少的。应用科学计量学与复杂网络分析的方法,选取网络的平均聚类系数、平均最短路径和平均度三个特征指标,建立科学知识网络的自相似模型,并对合作网络、共词网络与共被引网络的自相似性进行定性与定量的分析,从而验证了科学文献的网络拓扑结构的局部与整体具有自相似。

Self-similarity of scientific knowledge is the theoretical hypothesis of scientometrics. Especially the mappingknowledge is even based on the spatial self-similarity of scientific literatures. Therefore it is essential to investigate the selfsimilarityof scientific knowledge network. We applied scientometics and complex network analysis to study the self-similarity ofcooperative network, co-word network and co-citation network qualitatively and quantitatively, where select three characteristicindices which are the average clustering coefficient, average shortest path and the average degree of the network to establish a selfsimilaritymodel. In the resuilt, prove that the local network topology and global network topology of the scientific literature are selfsimilarity.