Knowledge acquisition with supervised ontology population

Ontology plays a crucial role in capturing and disseminating business information (e.g., products, services, relationships of businesses) for effective human computer interactions. However, manual construction of domain ontology is very labour intensive and time consuming. This paper illustrates a novel ontology population method for semi-automatic business knowledge acquisition from text. In particular, the proposed method is underpinned by the effective SVM-struct algorithm which treats ontology population as a sequence labelling problem. By automatically exploring taxonomy knowledge captured in domain ontology, our ontology population method can effectively classify objects to multiple categories. The initial experimental results show that the SVM-struct based ontology population method which utilizes taxonomy knowledge outperforms other traditional methods in a benchmark ontology population task.