A two-level approach to generate synthetic argumentation reports

Given a controversial issue, a major challenge in argumentmining is to organize the arguments which have been minedto generate a synthesis that is readable, synthetic enough andrelevant for various types of users. Based on the GenerativeLexicon (GL) Qualia structure, which is a kind of lexicaland knowledge repository, that we have enhanced in di↵erentmanners and associated with inferences and language pat-terns, we show how to construct a synthesis that outlines thetypical elements found in arguments. We propose a two-levelapproach: a synthesis of the arguments that have been minedand navigation facilities that allow to access the argumentcontents in order to get more details.