Arguer is an argumentation system that uses a schema-based approach to detecting and generating possible rebuttals to an argument. Schemata characterize the structure of arguments and are used to process input propositions (utteranc es). ARGUER maintains a model of the user's beliefs and uses this model to generate possible rebuttals to an utterance, or to detect whether a user utterance attacks a system utterance. It maintains separate models of beliefs for each participating agent; consequently, it needs not believe propositions that are believed by other agents. 

ARGUER does not assume a priori knowledge about attack and support relations between utterances advanced by the agents participating in a dialog. It establishes these relations dynamically, i.e, while the dialog is taking place. Moreover, the method used is symmetrical in that it can be used for interpretation or generation of arguments. This is important because the system can have the role of an observer or participant (or both). 

It processes propositions that may be part of an entire argument, incrementally, based on the current utterance advanced by a dialog participant, rather than generating or detecting complete arguments at once. In a dialog, input proposi tions (utterances) are often parts of an entire argument rather than the complete argument.