Extracting ontologies from World Wide Web via HTML tables
Minoru Yoshida, Kentaro Torisawa, Jun’ichi Tsujii · 2001
This paper describes a method to extract ontologies from tables in the World Wide Web (WWW). A table can be seen as a device to describe relating objects by attributevalue pairs. The attributes specify the important information that we need to know for identification and utilization of the described objects. This property is the same as the requirement on generic ontologies. So, by properly processing a wide range of tables, we can construct ontologies. We proposed an unsupervised method for this task. The method utilizes the EM algorithm and can be seen as an unsupervised learning method. The effectiveness of our method is confirmed by a series of experiments. 1