Automatic Learning of Discourse Relations in Swedish Using Cue Phrases

Abstract. This paper describes experiments to extract discourse relations holding between two text spans in Swedish. We considered three relation types: cause-explanation-evidence (CEV), contrast, and elaboration and we extracted word pairs eliciting these relations. We determined a list of Swedish cue phrases marking explicitly the relations and we learned the word pairs automatically from a corpus of 60 million words. We evaluated the method by building two-way classifiers and we obtained the results: Contrast vs. Other 67.9%, CEV vs. Other 57.7%, and Elaboration vs. Other 52.2%. The conclusion is that this technique, possibly with improvements or modifications, seems usable to capture discourse relations in Swedish. Key words: rhetorical relations, discourse relations, cue phrases, naïve Bayes classification.

• Type: Journal
• Year: 2010
• Where: TAL
• Authors: Stefan Karlsson, Pierre Nugues
• Keywords: Discourse, Natural Language Processing, Relation Types, TAL 2010, Text Spans