The Use of Ontology for Data Mining with Incomplete Data

Knowledge management has become one of the important topics in the database management field (Zhang and Zhao 2006). There have been two important research themes in knowledge management: data mining (Cunningham et al. 2006; Fayyad et al. 1996) and ontology (Green and Rosemann 2004; Kim 2002). Data mining is the process of trawling through data to find interesting patterns (Hand 1998). As such a process reveals previously unknown relationships among the data, data mining has become a widely used knowledge discovery technique (Brachman et al. 1996). On the other hand, ontology is a science that studies explicit formal specifications of the resources and relations among them in the domain (Gruber 1993). An ontology is a specification of a conceptualization (Gruber 1995), and intended for knowledge sharing among applications (Welty 2003). In the past few years, the two themes have become well-recognized substrate for research into knowledge management (Nilakanta et al. 2006; Li and Zhong 2006). Yet, potential benefits of joining the two themes have not been explored. Intuitively, the use of ontology for data mining can be beneficial for knowledge management in the following aspects.