基于关联数据的馆藏文物资源知识关联与智能问答研究

Research on Knowledge Association and Intelligent Question-and-Answer of Cultural Relic Resources Based on Linked-Data

【目的/意义】馆藏文物资源知识分散、异构、关联程度较低,本文提出一种馆藏文物资源知识关联的有效方法,便于关联化的知识服务的开展和馆藏文物信息资源的开发利用。【方法/过程】本文在分析馆藏文物信息资源关联和智能问答现状的基础上,提出基于关联数据的馆藏文物信息资源知识关联方法,构建了一个基于关联数据的知识关联模型。通过借助关联数据相关技术、利用SPARQL以及HTML对馆藏文物资源知识进行访问和查询。【结果/结论】将馆藏绘画类文物知识发布为关联数据形式后,馆藏文物资源知识得到充分的聚合,深度展现文物信息资源之间的关系,提高了不同需求的用户浏览不同粒度的文物知识、知识实体关系的效率。【创新/局限】从关联数据角度出发,提出基于多粒度语义查询的智能问答服务框架。下一步研究需要扩展馆藏文物资源的数据采集规模,对馆藏文物资源的深层次关联发现进行探索。

【Purpose/significance】The knowledge of cultural relic is scattered,heterogeneous,and low relevance.This paper proposes an effective method of knowledge association of cultural relics resources,which is convenient for the development of related knowledge services and the development and utilization of cultural relics information resources.【Method/process】Based on the analysis of the status quo of cultural relics information resource and Intelligent Question-and-Answer in collections,this paper proposes a method of knowledge relevance of cultural relics information resources based on Linked-data,and constructs a knowledge association model based on Linked-data.With the help of Linked-data technology,SPARQL and HTML to access and query the knowledge of cultural relics in the collection.【Result/conclusion】After publishing the knowledge of cultural relics in the collection of paintings in the form of Linked-data,the knowledge of cultural relics in the collection is fully aggregated,and the relationship between cultural relics and information resources is displayed in depth,and improved the efficiency of users with different needs in browsing cultural relics knowledge and knowledge entity relationships of different granularities.【Innovation/limitation】From the perspective of linked data,an intelligent question answering service framework based on multi-granular semantic query is proposed.The next step of research needs to expand the scale of data collection of cultural relics in the collection,and explore the deep-level association discovery of it.