Abstract Konstantopoulos
This is to present work on applying Inductive Logic Programming (ILP) to the task of NP chunking. A first attempt (presented in CLIN 1999 [1]) yielded modest results, but suggested ways of improving both the predictive accuracy of the induced theory, as well as its quality (i.e. compactness and intuitiveness.)The experiments presented here differ from the original ones of [1] in the size of the training data as well as the sophistication of the background knowledge made available to the ILP algorithm. The resulting theories are qualitatively analysed and their performance compared to that of other machine learning approaches to chunking. REFERENCE
[1] Stasinos Konstantopoulos, NP Chunking using ILP, Proc. of CLIN 1999, pp. 109-116, Utrecht, 2000
Last modified $Date: 2001/10/04 13:39:46 $ by Parlevink Webmaster