Automated parser construction

Application domain: Parser construction
Pointers: http://nlp.fi.muni.cz/projekty/gram_inf/
Data format: Prolog

The task addressed is learning a sequence of context-dependent parse actions from a given corpus of labelled derivation trees. To this end, the system GRIND combines two established methods of machine learning: transformation-based learning (TBL) and inductive logic programming (ILP). Being trained and tested on the corpus SUSANNE, GRIND reached the accuracy of 96% and the recall of 68%.

References

  1. Nepil M. Automated Parser Construction from a Treebank by means of TBL and ILP. In Proceedings of the Student Research Workshop at ACL/EACL 2001, Toulouse, France. Association for Computational Linguistics, 2001, pp. 19-24.

  2. Nepil M. Learning to Parse from a Treebank: Combining TBL and ILP In Rouveirol C. and Sebag M. (eds.): Proceedings of Eleventh International Conference on Inductive Logic Programming, ILP 2001, Strasbourg, France. Springer Verlag, 2001, LNAI 2157, pp. 179-192.

back to index