Extracting structured knowledge from Dutch legal texts: a rule-based approach

Abstract

Legal texts are difficult to interpret, and its interpretation depends on the knowledge and experience of

the legal expert. Formalising interpretations can improve transparency. However, creating formalisations

of legal texts is labour-intensive, and automatically creating them is still a challenge. Previous work

showed that rule-based systems have mixed success on Dutch legal texts. They use complex rule systems for specific cases, making them hard to compare. Because of the lack of analysis, the success of these methods is also unclear. In this paper, we propose a new rule-based architecture for detecting the different roles of Flint frames, a knowledge representation language which aims to be a generic and less task-dependent language. The rules in this architecture are based on Part-of-Speech tags and universal dependency tags. Our analysis shows that this combination yields more precise extraction of the roles of Flint frames than previous methods, and the use of universal dependency tags allows this method to also be applied to other languages. For further improvement we suggest extending the rules for extracting the recipient role, add rules for recognising complex relative clauses, and testing this framework on English legal texts.