摘要
[目的/意义]对现有面向技术路径识别的文本挖掘方法应用研究相关内容进行总结梳理,辨析当前局限与未来发展方向,为进一步完善该研究方法体系、提升技术路径识别与预测应用效果提供理论支持。[方法/过程]系统梳理了相关研究成果,从主要应用的文本挖掘技术类型视角总结了现有研究的内容与特点,分析了当前研究的局限性,探讨了未来发展方向。[结果/结论]当前面向技术路径识别的文本挖掘方法主要有词频分析、文本主题特征识别、语义增强、文本关联、文本聚类等;引文网络与文本挖掘相结合的方法应用广泛;多元融合方法是未来数据分析发展趋势。主要局限性:文本主题特征识别范围不够全面,语义关联定量化测定有待加强,主题关联数据融合方式单一化,路径预测时效性与动态性较弱。未来研究方向:提高文本特征识别准确性,探索语义化技术结构的路径预测方法,加强面向多源数据的多元关系融合研究,加强动态性与预测性。
[Purpose/significance]This study summarizes existing researches on text mining methods oriented to technical path identification,discriminates the current limitations and the future direction of development,and can provides theoretical support for further improving the research method system and the application effect of technology path recognition and prediction.[Method/process]This study also systematically presents relevant research results,summarizes the content and characteristics of the existing research from the perspective of text mining technology types,analyzes the limitations of the current research,and discusses the development direction of future.[Result/conclusion]At present,text mining methods oriented to technical path recognition mainly include word frequency analysis,text topic feature recognition,semantic enhancement,text association,text clustering,etc.The method of combining citation network and text mining is widely used.Multiple fusion is the trend of data analysis in the future.Main limitations:the scope of text subject feature recognition is not comprehensive enough;the quantitative measurement of semantic correlation needs to be strengthened;the fusion mode of topic related data is singular;the path prediction is weak in timeliness and dynamics.Future research directions:improving the accuracy of text feature recognition,exploring the path prediction method of semantic technical structure,strengthening the research on multi-relationship fusion for multi-source data,and enhancing the dynamics and predictability.
出处
《情报理论与实践》
CSSCI
北大核心
2020年第7期179-185,共7页
Information Studies:Theory & Application
基金
国家社会科学基金项目“技术创新路径识别与预测的多元关系融合方法研究”的成果,项目编号:18BTQ067。
关键词
技术路径识别
文本挖掘
技术演化
多元关系融合
technological path identification
text mining
technological evolution
data fusion of multi-relationships
