A methodology for domain ontology construction in information science

Domain ontologies have been widely acknowledged as fundamental in supporting knowledge-based applications. To reduce the bottleneck of knowledge acquisition in the construction of domain ontologies text mining techniques have been used. This paper proposes a methodology for building domain ontology for information science through extraction of knowledge from related literature and thesaurus. It involves collecting domain-related vocabularies, developing the domain concept hierarchy, and defining the properties and relationships between concepts. The core concepts in information science are selected by applying social network theory and the relationships between concepts are discovered by hierarchical clustering method. Our methodology can be applied to other domain ontologies, such as computer science and management science.