State of the ART: An Argument Reconstruction Tool

This paper describes the outcomes of a series of experiments in automated support for users that try to find and analyse arguments in natural language texts in the context of the FP7 project IMPACT. Manual extraction of arguments is a non-trivial task and requires extensive training and expertise. We investigated several possibilities to support this process by using natural language processing (NLP), from classifying pieces of text as either argumentative or non-argumentative to clustering answers to policy green paper questions in the hope that these clusters would contain similar arguments. Results are diverse, but also show that we cannot come a long way without an extensive pre-tagged corpus.