摘要
[目的/意义]知识单元作为人类控制和管理知识的基本单元,是科学文献内容的基本组成成分。通过对知识单元抽取技术和应用研究进行梳理与总结,探索知识单元相关研究未来发展方向,以便更好地实现对科学知识的发掘和利用。[方法/过程]采用文献调研法,通过人工对相关文献进行阅读和分析,从知识单元抽取和应用两个视角对科学文献中知识单元相关研究进行梳理和总结。[结果/结论]目前,知识单元抽取的准确性和智能性还有待于进一步提高,未来研究可关注新兴深度学习技术在知识挖掘中的应用;当前知识单元的应用主要包括科学知识图谱构建、学术评价和知识演化分析,未来研究可进一步探索知识单元更多的应用场景,包括基于引文内容的知识单元评价、基于知识图谱的学术文献创新性评价和基于知识单元的跨学科知识流动分析等。
[Purpose/significance]As the basic unit of human control and management,knowledge unit(KU)is the basic component in the content of the scientific paper.By sorting out and summarizing KU extraction technology and applied research,the future development direction of KU related research can be explored,so as to better realize the exploration and utilization of scientific knowledge.[Method/process]This paper adopts the literature research method,through manual reading and analysis of relevant literature,to sort out and summarize related research on KUs in scientific literature from two perspectives of KU extraction and application.[Result/conclusion]After sorting out and analyzing,the accuracy and automation of KUs extraction need to be improved,future research can focus on the application of emerging deep learning technologies in knowledge mining;the current application of KUs mainly includes scientific knowledge graph construction,academic evaluation and knowledge evolution analysis.Future research can further explore more application scenarios of KUs,including the evaluation of KUs based on citation content;the innovative evaluation of academic literature based on knowledge graph,and the analysis of interdisciplinary knowledge flow based on knowledge units.
出处
《情报理论与实践》
CSSCI
北大核心
2022年第12期195-207,共13页
Information Studies:Theory & Application
基金
国家社会科学基金重点项目“基于关联数据的学术文献内容语义发布及其应用研究”的成果之一,项目编号:17ATQ001。
关键词
知识抽取
知识单元
知识实体
科学文献
knowledge extraction
knowledge unit
knowledge entity
scientific literature