Finetuning NLG Through Experiments with Human Subjects: The Case of Vague Descriptions

Abstract

This discussion paper describes a sequence of experiments with human subjects aimed at finding out how an NLG system should choose between the different forms of a gradable adjective. This case study highlights some general questions that one faces when trying to base NLG systems on empirical evidence: one question is what task to set a subject so as to obtain the most useful information about that subject, another question has to do with differences between subjects. © Springer-Verlag 2004.