A Framework for Capturing Domain Knowledge via the Web
Chao Wang, Jie Lü, Guangquan Zhang · UTS ePRESS (University of Technology Sydney) · 2005
Domain knowledge can be formalized and represented by ontologies, which play an important role in the realization of the Semantic Web. However, since the acquisition of knowledge from certain domains usually requires deep involvement of qualified domain experts, construction of such ontologies is difficult and costly, even with the availability of dedicated languages and ontology editing tools. Some effect has been made to reduce this involvement by introducing a general paradigm of automatic domain knowledge learning from various sources. To make this paradigm more specific and practical, this paper proposes a framework for capturing domain knowledge through raw domain data available over the Web. This framework consists of three dedicated parts: data collection, pre-processing and mining, where mining part performs core task of the framework. Each part can be designed with specific optimized methods. The preliminary implementation of certain parts has shown it is able to capture the knowledge of electronic product taxonomy via the Web. © 2005. Chao Wang, Jie Lu & Guangquan Zhang.