We introduce a chunking system based on probability distribution approximation and decision trees. The system was tested on the large standard evaluation set for base noun phrases. While the training time of the system is remarkable short (about 3 hours), results are comparable with the best systems reported so far. We trained our system with different settings of POS tags and show how much chunking results depend on POS tag accuracy. © Springer-Verlag Berlin Heidelberg 2003.
CITATION STYLE
Lüdtke, D., & Sato, S. (2003). Fast base NP chunking with decision trees - Experiments on different POS tag settings. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2588, 136–147. https://doi.org/10.1007/3-540-36456-0_14
Mendeley helps you to discover research relevant for your work.