基于引文耦合和概念格的学科交叉知识结构探测

Detecting Interdisciplinary Knowledge Structure Based on Concept Lattice and Bibliographic Coupling

[目的/意义]提出基于引文耦合和概念格的学科交叉知识结构探测方法。[方法/过程]利用Sci^2工具获取情报学与计算机科学的学科交叉耦合文献,利用ConExp1.0工具构建两学科交叉知识概念格;通过概念格关联规则挖掘和层次聚类,将两学科的学科交叉知识结构细分为8个研究主题,揭示各研究主题的关联特征和对应的核心作者。[结果/结论]与传统的聚类树相比,该方法结合人类的认知规律描述概念与主题之间的关系,不仅具有较好的层次性可视化效果,而且意义更易于理解,具有更好的解释性。

[ Purpose/significance ] The interdisciplinary knowledge structure detecting method based on concept lat- tice and bibliographic coupling is proposed. [ Method/process] Firstly, it obtains the interdisciplinary coupling literature of information science and computer science using Sci2, and builds interdisciplinary knowledge concept lattice using ConExpl. 0. Then, the interdisciplinary knowledge structure is subdivided into eight major research themes through association rule mining and hierarchical clustering based on concept lattice. Finally, it reveals the correlation characteristics of each research topics and the corresponding core author of each theme.[Result/conclusion ] Compared with the traditional clustering tree for discovering knowledge structure, the method describes the relationship between the concept and theme combined with human cognitive law, and has good gradation visual effect for easier to understand and better explanation.