"My research looks at the development of computational linguistic models that automatically recognise how people debate with one another. Important aspects include the type of reasoning and the role of questions and answers. The main question is: Which models are most effective at identifying, classifying, and interpreting these patterns in large amounts of data? For one thing, this drives forward computational linguistic research, because many elements of a debate remain implicit and are thus not found on the linguistic surface. For another, this research has implications for theoretical models of communication and argumentation in as much as the patterns found in large amounts of data regularly undermine theoretically made assumptions. This interdisciplinarity informs the bulk of my work because the ultimate goal is to develop robust models that facilitate solid knowledge gains in both fields of research."
More on her research
In her research, Professor Hautli-Janisz studies computer linguistic modelling of discourse, rhetorical strategies, and argumentation in naturally spoken language. She develops and employs a broad range of natural language processing models, from rule-based and statistical approaches all the way to neural networks.
More on the research of Professor Annette Hautli-Janisz.