CaTeRS: Causal and Temporal Relation Scheme

  • Nasrin Mostafazadeh
  • Alyson Grealish
  • Nathanael Chambers
  • James Allen
  • Lucy Vanderwende

Proceedings of the 4th Event Workshop at NAACL HTL |

Learning commonsense causal and temporal relation between events is one of the major steps towards deeper language understanding. This is even more crucial for understanding stories and script learning. A prerequisite for learning scripts is a semantic framework which enables capturing rich event structures. In this paper we introduce a novel semantic annotation framework, called Causal and Temporal Relation Scheme (CaTeRS), which is unique in simultaneously capturing a comprehensive set of temporal and causal relations between events. By annotating a total of 1,600 sentences in the context of 320 five-sentence short stories sampled from ROCStories corpus, we demonstrate that these stories are indeed full of causal and temporal relations. Furthermore, we show that the CaTeRS annotation scheme enables high inter-annotator agreement for broad-coverage event entity annotation and moderate agreement on semantic link annotation.