Inductive learning of attribute path values in typed-unification grammars.

Published in Volume XIII, 2003, pages 105-126

Authors: Liviu CIORTUZ

Abstract

This work is aiming to show that inductive logic programming (ILP) is a suitable tool to learn linguistic phenomena in typed-unification grammars — also known as typed feature structure (TFS) grammars —, a class of attribute-value grammars increasingly used in natural language processing (NLP).

Bibtex

@article{sacscuza:ciortuz2003iloapvitg,
  title={Inductive learning of attribute path values in typed-unification grammars.},
  author={Liviu CIORTUZ},
  journal={Scientific Annals of Computer Science},
  volume={13},
  organization={``A.I. Cuza'' University, Iasi, Romania},
  year={2003},
  pages={105--126},
  publisher={``A.I. Cuza'' University Press}
}