Finding temporal and causal relations is crucial to understanding the semantic structure of a text. Since existing corpora provide no parallel temporal and causal annotations, we annotated 1000 conjoined event pairs, achiev
Steven Bethard, James H. Martin