Web mining for causal relations between events

This paper addresses the problem of automatic acquisition of causal relations between events. Since most of the previous researches rely on annotated corpus, the main challenge is the need for more generic methods to identify cause-effect pairs and to extract event-arguments (particularly the predicate, subject and object). Motivated by this background, we develop a three-phased approach that acquires causality from the Web. Firstly, we use explicit connective markers (such as "because") as linguistic cues to discover causal related events. Then, we extract the event-arguments based on local dependency parse trees of event expressions. In the final phase, we propose a statistical model to measure the potential causal relations. Experimental results demonstrate the effectiveness of our approach.