论文部分内容阅读
可视化同被引分析步骤主要包括选择数据源、确定分析的元素、计算同被引频次、对数据进行标准化处理、对数据进行降维和图示、对图形进行分析解释。这些步骤中的关键技术在于对获得的高维引文关系数据进行降维和图示,目前应用较多的主要有聚类分析、多维尺度分析、因子分析、自组织映射神经技术、寻径网络技术、最小生成树法、三角测量、力矢量布局算法等。其中寻径网络技术以其较好的图示效果得到较多的应用。
Visual citation analysis steps include selecting the data source, determining the elements of analysis, computing the same citation frequency, standardizing the data, performing data reduction and graphing, and analyzing and interpreting the graph. The key techniques in these steps are dimensionality reduction and graphing of the obtained high-dimensional citation relational data. At present, there are mainly applied clustering analysis, multidimensional scaling analysis, factor analysis, self-organizing mapping neural technology, routing network technology, Minimum spanning tree method, triangulation, force vector layout algorithm. Pathfinder network technology for its better pictorial effect get more applications.