Section: New Results

Identification of implicit discourse relations

Participant : Chloé Braud.

In collaboration with Pascal Denis (Magnet, Inria), we have developed a system for identifying “implicit” discourse relations (that is, relations that are not marked by a discourse connective) [33] . Given the little amount of available annotated data for this task, our system also resorts to additional automatically labeled data wherein unambiguous connectives have been suppressed and used as relation labels, a method introduced by Marcu and Echihabi (2002). As shown by Sporleder and Lascarides (2008) for English, this approach doesn't generalize well to implicit relations as annotated by humans. We have shown that the same conclusion applies to French due to important distribution differences between the two types of data. In consequence, we propose various simple methods, all inspired from work on domain adaptation, with the aim of better combining annotated data and artificial data. We have evaluated these methods through various experiments carried out on the ANNODIS corpus: our best system reaches a labeling accuracy of 45.6%, corresponding to a 5.9% significant gain over a system solely trained on manually labeled data.