BDBComp
Parceria:
SBC
Summarizing RST trees focusing on referential chains: A case study

Eloize Rossi Marques SenoLucia Helena Machado Rino

In this paper we introduce RHeSumaRST, a heuristics based system that aims at pruning an RST tree in order to yield its corresponding summary. Those heuristics focus especially on identifying when a discursive segment that embed an anaphoric term is chosen to compose the summary, in which case the discursive segment that embed its antecedent must also be chosen. To both address referentiality and rhetorical structuring, Veins Theory is added to RST Theory in order to drive pruning. A preliminary case study is presented that assesses pruning of RST trees resulting from discourse-analyzing texts written in Brazilian Portuguese.

http://www.lbd.dcc.ufmg.br/colecoes/til/2005/004.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD