摘要
应用拉普拉斯谱分析方法研究科学论文的词同现网络,从拉普拉斯谱逆序分布、拉普拉斯谱密度、拉普拉斯谱特征值极值三个角度对比分析真伪科学论文词同现网络的拉普拉斯谱特征。通过对比分析发现,真伪论文词同现网络的拉普拉斯谱特征存在明显差别,可以利用这些差别对真伪论文进行甄别。
This paper studies the Laplacian spectrum of scientific papers＇ word co - occurrence network from three aspects： the Laplacian spectrum rank, the Laplace spectral density, and the Laplacian extremal eigenvalues. Through comparative analysis, the Laplacian spectral features of scientific papers＇ word co - occurrence network are obviously different, and these differences can be used for discriminating the authenticity of scientific paper.
出处
《现代图书情报技术》
CSSCI
北大核心
2010年第5期58-65,共8页
New Technology of Library and Information Service
关键词
复杂网络
词同现网络
拉普拉斯谱
拉普拉斯谱密度
Complex network Word co- occurrence network Laplacian spectrum Laplace spectral density
