Paper: Automatic Detection of Causal Relations in German Multilogs

ACL ID W14-0703
Title Automatic Detection of Causal Relations in German Multilogs
Venue Workshop on Computational Approaches to Causality in Language
Year 2014

This paper introduces a linguistically- motivated, rule-based annotation system for causal discourse relations in transcripts of spoken multilogs in German. The over- all aim is an automatic means of determin- ing the degree of justification provided by a speaker in the delivery of an argument in a multiparty discussion. The system comprises of two parts: A disambiguation module which differentiates causal con- nectors from their other senses, and a dis- course relation annotation system which marks the spans of text that constitute the reason and the result/conclusion expressed by the causal relation. The system is eval- uated against a gold standard of German transcribed spoken dialogue. The results show that our system performs reliably well with respect to both tasks.