Studying EU legislation on a large scale is difficult because of its considerable amount and complex nature. In 2012 EU legislation and quasi-egislation comprised approximately 30,000 documents of more than 100,000 pages which often contains multidimensional semantic information. However, recent progress within computer science with regard to machine learning carries great potential for developing advanced automatic tools for analyzing data which will make it possible to discover latent communication patterns invisible to the human eye. This paper has an interdisciplinary aim in using cutting edge techniques from linguistic content-based information retrieval to answer a key research question within Law and Political Science: how often do the Commission, Council of Ministers, and European Parliament refer to judgments of the European Court of Justice on the institutional balance of power and the “inter-institutional agreements”?
