Research on Domain Ontology Construction in Digital Library

2018 International Conference on Advanced Electronic Materials, Computers and Materials Engineering

材料科学;无线电电子学;计算机科学

Li, Xiao-Fei^1

Nangchang Normal University, Nangchang, China^1

关键词: Domain consistency;    Domain ontology constructions;    Hierarchical clustering methods;    Information entropy;    Knowledge organization;    Learning behavior;    Log-likelihood ratio algorithm;    Relationship between concepts;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/439/3/032005/pdf
DOI  :  10.1088/1757-899X/439/3/032005

来源: IOP

【 摘 要 】

Learning resource navigation is a major problem that restricts learners' learning behavior. This paper proposed semantic approach based on ontology, which is helpful to solve the knowledge organization problem of digital library. The corpus was collected according to the 22 major categories of the Chinese Library Classification. It was divided into domain corpus and balanced corpus. Based on the knowledge base, information entropy and variance, linguistics, combined word length ratio, domain correlation and domain consistency, the domain concept was extracted. The cohesion hierarchical clustering method was used to extract the classification relationship between concepts. The log-likelihood ratio algorithm was used to extract the non-classification relations of domain concepts.