Ontologies are an essential component in Information Systems since they enable knowledge re-use and sharing in a formal, homogeneous and unambiguous way. A domain ontology captures knowledge in a static way, as it is snapshot of knowledge from a particular point of view in a specific time-period. However, in open and dynamic settings, where knowledge changes and evolves, ontology maintenance methods are required to keep knowledge up-to-date. In this chapter we tackle the problem of ontology maintenance as an ontology population problem of the evolving ontologies proposing an incremental ontology population methodology that exploits machine learning techniques and is enforced with a bootstrapping technique in order to tackle large scale problems. The methodology is enriched with fine-tuning methods towards improving the quality and the number of the discovered instances. Finally, experimental results are shown, which prove the applicability and effectiveness of the proposed methodology.
CITATION STYLE
Valarakos, A. G., Vouros, G., & Spyropoulos, C. (2007). Machine Learning-Based Maintenance of Domain-Specific Application Ontologies. In Ontologies (pp. 339–372). Springer US. https://doi.org/10.1007/978-0-387-37022-4_12
Mendeley helps you to discover research relevant for your work.