情报科学

• • 上一篇    下一篇

基于知识元的文物图像资源细粒度知识表示方法研究

  

  1. 华中师范大学信息管理学院

A Fine-grained Knowledge Representation Method of Cultural Heritage Image Resources Based on Knowledge Element

  1. School of Information Management, Central China Normal University, Wuhan, 430079, China

摘要:

【目的/意义】文物图像资源日益成为数字人文基础设施建设的重要内容,为了克服存在于图像资源内容与形式上的“语义鸿沟”对其开发利用产生的消极影响,有必要面向文物图像资源底层视觉特征与高层语义特征的精细化映射与细粒度知识表示进行相关研究。【方法/过程】本文在分析文物图像资源知识表示需求与表示策略的基础上,提出了基于知识元构造的文物图像资源细粒度知识表示模型。在模型设计的基础上,以著名绘画文物《历代帝王图卷》为例,阐述了面向文物图像资源细粒度表示需求进行知识元提取、构造与数据发布的具体流程。【结果/结论】实验结果表明,本文提出的基于知识元的文物图像资源细粒度知识表示方法能够在图像底层视觉特征与高层语义特征之间建立有效的语义映射关系,并通过与外部知识库的数据关联实现与开放关联数据网络的深度融合。【创新/局限】本文从知识元角度出发,提出了文物图像资源的细粒度知识表示方法,在未来的研究中还需对文物图像知识元的自动提取以及基于知识元的图像资源知识发现方法进行更深入地探索。

关键词:

文物图像资源, 知识元, 细粒度, 知识表示, 本体, 数字人文

Abstract:

 [Purpose/Significance] Cultural heritage image resources are increasingly becoming an important part of the infrastructure construction of digital humanities. In order to overcome the negative impact of the semantic gap between the content and form of image resources on their development and utilization, it is necessary to study the fine mapping and fine-grained knowledge representation of the low-level visual features and high-level semantic features of cultural heritages image resources.[Method/Process] Based on the analysis of the requirements and strategies of knowledge representation in the field of cultural heritage image resources, this paper proposes a fine-grained knowledge representation model of cultural heritage image resources upon knowledge element. On the basis of model building, this paper take the famous painting heritage-The Thirteen Emperors as an example, elaborates the specific process of knowledge element’s extraction, construction and release so that meeting the fine-grained representation requirements of cultural heritage image resources. [Result/Conclusion] The experimental results show that the fine-grained knowledge representation method based on knowledge element can establish an effective semantic mapping relationship between the low-level visual features and the high-level semantic features of the image, and realize the deep integration with the open relational data network through data association with external knowledge base. [Innovation/Limitation] This paper designs and demonstrates the fine-grained knowledge representation method of cultural heritage image resources based on knowledge element. In subsequent studies, we need to further explore the automatic extraction method of cultural heritage knowledge elements, and the knowledge discovery of cultural heritage image resources upon knowledge elements.

Key words:

cultural heritage image resources, knowledge element, fine-grained; knowledge representation, ontology, digital Humanities