情报科学 ›› 2021, Vol. 39 ›› Issue (4): 9-14.

• 专论 • 上一篇    下一篇

基于主题识别的文物信息资源知识发现方法研究

  

  • 出版日期:2021-04-01 发布日期:2021-04-09

  • Online:2021-04-01 Published:2021-04-09

摘要:

【目的/意义】文物信息资源专业性、复杂性的特征成为用户获取文物知识的一大障碍,提出一种能够有效
进行文物信息资源知识发现的方法对于中华文化的弘扬与传承以及文物有关研究资源的获取有着十分重要的意
义。【方法/过程】文章从文物信息资源研究对象的角度出发,首先利用LDA主题模型对信息资源文本中的隐含主题
进行挖掘。随后以研究对象作为检索入口,使用SPARQL查询语言在外部知识库中获取与文物有关的知识三元
组。最后将文本主题与知识三元组进行耦合,实现了针对文物信息资源的知识发现。【结果/结论】文章方法在两种
类别的文物信息资源中实现了较好的知识发现效果,为文物信息资源的知识发现提供了一种可行的发现方法。【创
新/局限】外部知识库中庞大的知识储备为文章方法从开放领域获取信息资源进行知识发现提供了支持,同时由于
外部知识库中包含的文物知识有限,文章方法较人工进行的知识发现结果尚有一定差距。

Abstract:

【Purpose/significance】The specialty and complexity of cultural relic information resources have become a major obstacle
for users to acquire cultural relic knowledge. The proposal of effective method for knowledge discovery of cultural heritage information
resources is important for the promotion and transmission of Chinese culture and the acquisition of research resources related to cultur⁃
al heritage.【Method/process】From the perspective in cultural relics information resources, the article firstly mines the main topics in
the text of information resources using LDA topic model. Then, using the research object as the retrieval portal, the knowledge triad re⁃
lated to cultural relics is obtained in the knowledge base with SPARQL query language. Finally, topics are coupled with the knowledge
in cultural relics information resources.【Result/conclusion】The of method article complete knowledge discovery in two categories of
heritage information resources, and provides a feasible discovery method for knowledge discovery of heritage information resources.【In⁃
novation/limitation】The huge knowledge reserve in knowledge base provide support for the article method to obtain information re⁃
sources from the open domain. The article method still has some gaps in the knowledge discovery results compared with those conduct⁃
ed manually due to the limitation in knowledge base.