| 作者混合共引网络对知识图谱绘制的改进研究 |
| |
| 引用本文： | 黄文彬,蒙汪阳,步一.作者混合共引网络对知识图谱绘制的改进研究[J].图书情报工作,2017,61(3):118-124. |
| |
| 作者姓名： | 黄文彬 蒙汪阳 步一 |
| |
| 作者单位： | 1. 北京大学信息管理系 北京 100871;2. 美国印第安纳大学信息学与计算机学院 印第安纳布鲁明顿 47408 |
| |
| 基金项目： | 本文系中国科技信息研究所系所合作项目研究成果之一。 |
| |
| 摘 要： | 目的/意义]作者共引网络分析（ACNA）是文献计量学中的重要分析方法，旨在通过寻找学术文献集合中作者之间的共引关系绘制出特定领域的知识图谱，进而指导科学研究。然而，ACNA的一个缺陷是其原始矩阵输入信息量过小。本文通过提出作者混合共引网络（HACNA），绘制更为精确的科学知识图谱。方法/过程]鉴于不同种类的学术网络能为绘制知识图谱提供不同维度的信息，提高知识图谱绘制的精确性，本文以合著网络和引用网络为例，结合其他种类的学术网络在ACNA基础上进行精确科学知识图谱的绘制。结果/结论]实证研究结果显示，与ACNA相比，HACNA绘制出的知识图谱在聚类过程中能够使得同类作者更为聚拢、不同类作者更为分散，从而提高了聚类效果和可视化程度。同时，HACNA绘制出的知识图谱还能够挖掘出更多细节。
|
| 关 键 词： | 作者共引网络 共引网络 合著网络 引文网络 文献计量学 |
| 收稿时间： | 2016-09-28 |
Improvements on Mapping Knowledge Domains by Using Hybrid Author Co-citation Network |
| |
| Huang Wenbin,Meng Wangyang,Bu Yi.Improvements on Mapping Knowledge Domains by Using Hybrid Author Co-citation Network[J].Library and Information Service,2017,61(3):118-124. |
| |
| Authors: | Huang Wenbin Meng Wangyang Bu Yi |
| |
| Institution: | 1. Department of Information Management, Peking University, Beijing 100871;2. School of Informatics and Computing, Indiana University, Bloomington, Indiana 47408 |
| |
| Abstract: | Purpose/significance] Author co-citation network analysis (ACNA) is an important method in bibliometrics which aims to map knowledge domains and to guide scientific research by considering cocitation relationships between author pairs in dataset (see reference #1). However, it is criticized that the amount of information in raw co-citation matrix is limited in ACNA. This paper proposes hybrid author co-citation network (HACNA) in order to map knowledge domains more accurately.Method/process] Because the accuracy of mapping knowledge domains could be improved due to different perspectives provided by multiple scholarly networks, this paper combines other types of scholarly networks (taking coauthorship networks and citation networks as examples) into ACNA to show more accurate knowledge domain maps, which is call hybrid author co-citation network analysis (HACNA).Result/conclusion] Results show that compared with ACNA, HACNA makes authors in the same category closer and authors in different categories farther in knowledge domain maps clustering, and thus promotes the clustering performance and visualization. Moreover, it is able to mine more details. |
| |
| Keywords: | author co-citation network co-citation network coauthorship network citation network bibliometrics |
|
| 点击此处可从《图书情报工作》浏览原始摘要信息 |
| 点击此处可从《图书情报工作》下载免费的PDF全文 |
