基于关联数据的敦煌遗书图像知识组织模式的构建

Construction of Knowledge Organization Model of Dunhuang Documents Images Based on Linked Data

敦煌遗书图像蕴含丰富的文化内涵,对于研究中国古代社会历史、宗教与美术具有重要意义,但传统单一线性的图像检索方式不利于敦煌遗书图像隐性知识的挖掘,影响知识发现的深度与广度。而关联数据能够连接多源异构资源,实现多种资源的语义互联,既能促进管理标准化与规范化,又有利于提升图像内容的深入整合,同时,将关联数据应用于敦煌遗书图像在理论、实践与技术上都具备可行性。为此,本文针对敦煌遗书图像的物理特征与内容语义特征构建敦煌遗书图像层次模型,使用元数据描述后将这些元数据进行关联;同时,本文基于关联数据设计敦煌遗书图像知识关联的组织模式,其自底向上分为数据收集层、语义描述层、数据关联层与知识应用层四层,旨在改善图像检索效果并利于敦煌遗书图像的知识发现与智能查询。

The Dunhuang documents images contain rich cultural connotations,which are of great significance to the study of ancient Chinese social history,religion and arts.However,the traditional single linear image retrieval methods are not conducive to the excavation of the hidden knowledge of Dunhuang documents images,influencing the depth and breadth of knowledge discovery.The linked data can link multiple heterogeneous resources and realize semantic interconnection of multiple resources,which can not only promote management standardization,but also promote the deep integration of images content.At the same time,it is feasible to apply the linked data to the Dunhuang documents images in theory,practice and technology.To this end,this paper constructs a hierarchical model of Dunhuang document images according to its physical and semantic features,and uses metadata to describe and correlate these metadata.This paper designs a knowledge association organization model based on linked data,which is composed of four layers from bottom to top:data collection layer,semantic description layer,data linked layer and knowledge application layer,in order to improve the effect of image retrieval and benefit the knowledge discovery and intelligent query of Dunhuang documents images.