Multi-document summarization using discourse models.

Automatic multi-document summarization aims at reducing the size of texts while preserving the important content. In this paper, we propose some methods for automatic summarization based on two semantic discourse models: Rhetorical Structure Theory (RST) and Cross-document Structure Theory (CST). These models are chosen in order to properly address the relevance of information, multi-document phenomena and subtopical distribution in the source texts. The results show that using semantic discourse knowledge for content selection improve the informativeness of automatic summaries.