Journal of Mathematical Sciences: Advances and ApplicationsAuthor's: Songil Cha, Kyongjin Sok, Changho Kim, Kinam Sin and Unsok Ryu
Pages: [59] - [89]
Received Date: May 20, 2019
Submitted by:
DOI: http://dx.doi.org/10.18642/jmsaa_7100122065
Ontology has become a fundamental and critical component for developing different applications on the Web. With the phenomenal growth of the Web resources, to generate automatically ontologies by using existing documents on the Web has gotten more and more attention. Forensic medicine ontology is used to narrate and reason the knowledge of a forensic domain. Previous studies for constructing forensic medicine ontologies have not automatically constructed by using all the semantic features of the Web documents. Hereby, it is difficult to rapidly construct an instance and property elements of ontology from the Web documents that are increasing daily. This paper focuses on the ontology generation method from forensic inquest documents structured on the Web. In order to generate most ontology elements for reasoning, we discuss how to generate hierarchical relationships, non-hierarchical relationships, and property elements based on structural and semantic characteristics of forensic inquest report. An empirical investigation of our method has proved the effectiveness of forensic medicine ontology generation. Experimental results show that our ontology generation method can generate with high quality rapidly and correctly important instances and properties of a forensic domain.
forensic medicine ontology, inquest report, Web tables, hierarchical relationship, class-instance.
