ASSESSING THE PLAUSIBILITY OF INFERENCE BASED ON AUTOMATED CONSTRUCTION OF CAUSAL NETWORKS USING WEB-MINING

Abstract

Causal networks have been utilized as a tool for structuring complex social phenomena systematically and visually. However, in many cases creating a causal network necessitates a great deal of interpretive works by analyzers. In this study, we have developed a new method for automating the creation of causal networks by utilizing text data on the Web. Examples of its application are illustrated using real cases. It is argued that the proposed method provides ways for verifying the validity of daily life policy discourse through existing causal statements.