论文部分内容阅读
【目的】利用LDA主题模型探索分类视角下的主题提取与分布状态,全面深入地揭示学科知识结构和热点。【方法】以国内知识流领域为研究对象,选取CNKI和万方数据库中知识流相关文献为数据源,利用中图分类号将知识流研究的文献分成11个学科,借助LDA主题模型分别对这11个学科所包含的文献进行主题抽取,挖掘出不同学科中所包含的20个热点主题。【结果】分析20个热点主题,得到11个不同学科的热点主题内容及其所揭示的知识点。【局限】该方法没有同其他学科主题挖掘方法进行对比,分析得到的国内知识流研究热点也没有同已有相关文献中分析出的该领域热点进行对照。【结论】该方法能够更全面和深入地挖掘学科知识结构和研究热点。
【Objective】 The theme of LDA is explored to explore the subject extraction and distribution under the perspective of classification, revealing the subject knowledge structure and hot spots thoroughly and deeply. 【Method】 Taking the field of knowledge flow in our country as the research object, the literature about knowledge flow in CNKI and Wanfang database was selected as the data source, and the literature of knowledge flow research was divided into 11 disciplines by using the Chinese Library Classification The topics included in the 11 subjects were extracted by topic, and the 20 hot topics included in different disciplines were excavated. 【Result】 Twenty hot topics were analyzed, and the hot topics content of 11 different subjects and their revealed knowledge points were obtained. [Limitations] This method has not been compared with the topic mining methods of other disciplines. The analysis of the domestic knowledge flow research hot spots has not been compared with the hot spots in the field that have been analyzed in the existing relevant literature. 【Conclusion】 This method can excavate the subject knowledge structure and research hot spot more comprehensively and deeply.