Learning Systems Engineering Domain Ontologies from Text Documents
Lan Yang,Kathryn Cormican,Ming Yu
DOI: https://doi.org/10.1109/isse46696.2019.8984550
2019-01-01
Abstract:Ontologies are playing an increasingly important role in knowledge management, and their functions have been appreciated and exploited by a broad range of communities, including systems engineering researchers and practitioners. Encompassing domain-related vocabularies, concepts, concept hierarchy, along with the properties and relationships, domain ontologies are becoming a promising medium for knowledge sharing and exchange. With the emergence of the semantic web and big data, learning domain ontologies from text is becoming a cutting-edge technique as it is an automatic process of deriving ontological knowledge. Specifically, a set of representative concepts and semantic relations can be rapidly derived from unstructured text documents in a hierarchical structure to model a domain. In this paper, we aim at exploiting the ontology learning approach to extract a domain ontology from systems engineering handbooks. An approach is proposed for learning terms, concepts, taxonomic and non-taxonomic relations. By incorporating both linguistic-based and statistical-based natural language processing techniques, we realized an automatic detection of complex domain terms and conceptualized the systems engineering body of knowledge in a semantic fashion. To evaluate the proposed approach, a case study is conducted, wherein the hybrid approach is applied with template-driven and machine learning algorithms. The result shows that the proposed approach has a robust performance in decreasing ontology development costs. This paper contributes to a good starting point for learning systems engineering ontologies to enhance knowledge acquisition and management.