Approximating theoretical linguistics classification in real data: the case of German nach particle verbs

摘要

Testing a theory against real world data can sometimes be helpful in figuring out the shortcomings of your current theory. In this paper, we test a theory about the syntax-semantics interface of German nach-particle verbs against data from a web corpus in order to see if we can use our automatic NLP machinery to corroborate the predictions of the theory We use state-of-the-art parsers to automatically annotate our data with the features predicted by the theory and then apply a standard clustering approach to approximate the nach-particle verb classes of the theory. The results of our experiment not only help us highlighting the more problematic parts of the theory but also teach us about the strengths and weaknesses of our automatic analysis tools.