What Kinds and Amounts of Causal Knowledge Can Be Acquired from Text by Using Connective Markers as Clues?

This paper reports the results of our ongoing research into the automatic acquisition of causal knowledge. We created a new typology for expressing the causal relations — cause, effect , precond(ition) and means — based mainly on the volitionality of the related events. From our experiments using the Japanese resultative connective “tame”, we achieved 80% recall with over 95% precision for the cause, precond and means relations, and 30% recall with 90% precision for the effect relation. The results indicate that over 27,000 instances of causal relations can be acquired from one year of Japanese newspaper articles.