作者三重耦合分析在知识图谱绘制中的应用研究
发布时间：2019-02-22 20:34
【摘要】：[目的/意义]提出三重耦合概念,以期通过改变传统耦合的作者频次计算方法,改进因偶然因素产生的过耦合现象,提高领域知识谱图绘制的准确度。[方法/过程]将原始矩阵构建从二重耦合计数改进为三重耦合计数,转化为相关矩阵后,对三维矩阵进行降维处理,通过Gephi软件绘制科学知识图谱并进行数据揭示与分析。[结果/结论]实证研究结果显示,三重耦合一方面保留了二重耦合的领域分析能力,另一方面提高了聚类结果的准确性,更为有效地进行作者可视化分析,有利于领域图谱绘制和子领域发现,挖掘出科学共同体的更多细节。
[Abstract]:[objective / significance] the concept of triple coupling is put forward in order to improve the over-coupling phenomenon caused by accidental factors and improve the accuracy of domain knowledge spectrum drawing by changing the traditional method of calculating the frequency of authors. [method / process] the original matrix construction was improved from double coupling count to triple coupling count, and then transformed into correlation matrix, the three-dimensional matrix was reduced dimension, and the scientific knowledge map was drawn by Gephi software, and the data was revealed and analyzed. [results / conclusions] empirical results show that triple coupling retains the domain analysis capability of double coupling on the one hand, and improves the accuracy of clustering results on the other hand, and makes the author visualize analysis more effectively. Facilitate domain mapping and subdomain discovery, mining out more details of the scientific community.
【作者单位】： 北京大学信息管理系;印第安纳大学信息学与计算机学院;
【分类号】：G353.1
本文编号：2428597
[Abstract]:[objective / significance] the concept of triple coupling is put forward in order to improve the over-coupling phenomenon caused by accidental factors and improve the accuracy of domain knowledge spectrum drawing by changing the traditional method of calculating the frequency of authors. [method / process] the original matrix construction was improved from double coupling count to triple coupling count, and then transformed into correlation matrix, the three-dimensional matrix was reduced dimension, and the scientific knowledge map was drawn by Gephi software, and the data was revealed and analyzed. [results / conclusions] empirical results show that triple coupling retains the domain analysis capability of double coupling on the one hand, and improves the accuracy of clustering results on the other hand, and makes the author visualize analysis more effectively. Facilitate domain mapping and subdomain discovery, mining out more details of the scientific community.
【作者单位】： 北京大学信息管理系;印第安纳大学信息学与计算机学院;
【分类号】：G353.1
【相似文献】
相关期刊论文 前1条
1 胡风娥;;图书馆三重和谐之构建[J];江西科技师范大学学报;2012年05期
，本文编号：2428597
本文链接：https://www.wllwen.com/tushudanganlunwen/2428597.html
