基于高频关键词的离群点监测与异类知识研究——从文献分析视角

发布时间：2018-10-16 10:33

【摘要】：[目的/意义]为了从海量数据中发现不易被发现的有价值的异类知识,避免知识的流失,从系统视角来对异类知识和离群点进行了探索。[方法/过程]首先对异类知识和离群点检测研究进行梳理。其次采用SNA,数据源采用中国知网和EBSCO为基础收录相关文献信息,利用文献题录信息分析软件BIBEXCEL构建共词矩阵,使用UCINET绘制了共现图,对离群点的高频关键词共现性和高产作者合著情况进行分析。并提供了可视化的图形来揭示研究该领域的发展趋势和现状。[结果/结论]研究结果表明现阶段对于异类知识的研究主要聚焦于outlier的挖掘和算法的提出、改进方面,而对异类知识管理机理方面的相关研究很少,不仅缺乏足够重视而且没有进行深入研究的探索。
[Abstract]:[objective / significance] in order to find the valuable and unusual knowledge from the massive data and avoid the loss of knowledge, this paper explores the different knowledge and outliers from a systematic perspective. Methods / process: firstly, the research on heterogeneity knowledge and outlier detection was combed. Secondly, the SNA, data source is used to collect the relevant literature information on the basis of China knowledge net and EBSCO, and the co-word matrix is constructed by using the document title information analysis software BIBEXCEL, and the co-occurrence map is drawn by UCINET. The cooccurrence of high frequency keywords in outliers and co-authorship of high-yield authors were analyzed. Visual graphics are also provided to reveal the development trend and current situation of this field. [results / conclusion] the results show that the current research on heterogeneous knowledge is mainly focused on the mining of outlier and the presentation and improvement of algorithms, but there is little research on the mechanism of heterogeneous knowledge management. Not only lack of enough attention, but also do not carry out in-depth research.
【作者单位】：武汉纺织大学管理学院;
【基金】：国家自然科学基金项目“大数据情景的outlier分析与异类知识管理模式研究”(编号:71571139) 湖北省高等学校人文社会科学重点研究基地-企业决策支持研究中心重大项目支持
【分类号】：TP311.13

【相似文献】