Rule-based information extraction: Advantages, limitations, and perspectives

With the clear advantages and capabilities of neural networks and machine learning in general, rule-based approaches and their usage and potential for information extraction (IE) are hardly addressed by academic research. This is counter-intuitive since recent studies have shown that a majority of large vendors of information extraction software, including IBM, SAP, and Microsoft, are using this technology predominantly within their offered products and services. This paper summarizes the advantages and limitations of rule-based IE and discusses its role for legal information retrieval. In addition, several different perspectives for academic research and industrial usage are illustrated.