PhD opportunities

A hybrid approach using formal rules and machine learning for the automatic understanding of the structure of the law and its evolution

Thesis proposal

Area of expertiseInformatique temps réel, robotique et automatique - Fontainebleau
Doctoral SchoolISMME - Ingénierie des Systèmes, Matériaux, Mécanique, Énergétique
SupervisorHERMANT Olivier
Co-supervisorSILBER Georges-André
Research unitMathématiques et Systèmes
KeywordsLaw, Machine Learning
AbstractThe objective of this thesis is to contribute to improve the intelligibility of the law for citizens and lawyers, by means of a graph representing the structure of the law and its evolution, obtained through the automatic understanding of legal texts. The law, described in a 'semi-formal' natural language, will have to be translated into a formal language describing the construction of this graph. The central problem of this translation is to automate the consolidation of legal texts over time by transforming the modification instructions contained in the modifying texts into modification programs.
ProfileNLP, machine learning, regular expressions, grammars, Python. The candidate must hold an engineering degree or a master of science. Proficiency in french and english, scientific writing in particular, is required.
FundingConcours pour un contrat doctoral