情报科学 ›› 2023, Vol. 41 ›› Issue (8): 134-138.

• 业务研究 • 上一篇    下一篇

科技文献的多粒度知识组织研究

  

  • 出版日期:2023-08-01 发布日期:2023-09-19

  • Online:2023-08-01 Published:2023-09-19

摘要: 【 目的/意义】基于粒度理论构建多粒度知识组织模型,旨在挖掘多粒度知识单元,并对其进行合理组织,满 足用户多粒度的知识需求。【方法/过程】首先,本文基于网络拓扑结构,构建了多粒度知识组织框架;其次,设计科 技文献多粒度知识组织模型,探究适用于不同粒度知识单元的知识表示方法和知识组织模型;最后,提出多粒度知 识组织的统一框架和体系化方法。【结果/结论】本文所构建的多粒度知识组织模型不仅能够很好地描述知识单元 的特征,还原多粒度知识单元之间的内在联系,还能够为用户提供多粒度的科学知识。【创新/局限】本文构架的科 技文献多粒度知识组织模型为多粒度科学知识的结构化表示和科学知识的横向、纵向关联提供了理论框架,本文 提出的科技文献多粒度知识组织流程为科技文献的多粒度挖掘和语义组织提供了相应的体系化方法,基于二者构 建多粒度知识组织图谱,不仅能够实现知识的多维度语义组织,还能够满足用户个性化、多层次科学知识需求。因 此,在未来的研究中,将不断完善多粒度知识组织模型,并建立面向科技文献的多粒度知识图谱。

Abstract:

【 Purpose/significance】 To construct a multi-granularity knowledge organization model based on granularity theory, aiming at mining multi-granularity knowledge units and organizing them rationally to meet users' multi-granularity knowledge needs.
【Method/process】 Firstly, this paper constructs a multi-granularity knowledge organization framework based on a network topology; secondly, it designs a multi-granularity knowledge organization model for science and technology literature, and explores the knowl⁃ edge representation methods and knowledge organization models applicable to multi-granularity knowledge units; finally, it proposes a unified framework and systematic method for multi-granularity knowledge organization. 【Result/conclusion】 The multi-granularity knowledge organization model constructed in this paper not only describes the characteristics of knowledge units well and restores the intrinsic connections among multi-granularity knowledge units, but also provides users with multi-granularity scientific knowledge.
【Innovation/limitation】 This paper constructed the multi-grained knowledge organization model of scientific and technical literature, which provides a theoretical framework for the structured representation of multi-grained scientific knowledge and the horizontal and vertical association of scientific knowledge. This paper proposed the multi-grained knowledge organization process of scientific and technical literature, which provides a corresponding systematic method for the multi-grained mining and semantic organization of sci⁃ entific and technical literature, based on the both, the multi-grained knowledge organization mapping can not only realize the multidimensional semantic organization of knowledge, but also meet the personalized and multi-level scientific knowledge requirements of users. Therefore, in the future research, the multi-granularity knowledge organization model will be continued and multi-granularity knowledge graphs for scientific and technical documents will be established.