The (semi-)automated integration of new information into a data model is a functionality which is required in cases when input documents are extensive and therefore a manual integration difficult or even impossible. We proposed an ontology learning procedure combining information acquisition from structured resources, such as WordNet or DBpedia, and unstructured resources using text mining techniques based on an evaluation of lexico-syntactic patterns. This approach offers a robust way, how to integrate even previously unknown information disregarding target application or domain. The proposed solution was implemented in the form of semi-automatic ontology learning tool used for integration of Excel document containing spare part records and Ford Supply Chain Ontology.
展开▼
