A General Method for Learning Idiosyncratic Grammars
Abstract
Natural language interfaces have fallen short of their goal of providing full freedom of expression in human-computer interactions largely because significant increases in the coverage of the grammar are usually accompanied by intolerable decreases in system performance. We demonstrate a method for overcoming this barrier called adaptive parsing, a technique in which the system accommodates the user by dynamically growing its grammar to acquire her preferred forms of expression and recognize them directly in the future. The specific learning method in an implemented adaptive parser is discussed in detail and shown to be adequate to acquire eight idiosyncratic grammars with good coverage and good performance.
Cite
Text
Lehman. "A General Method for Learning Idiosyncratic Grammars." International Conference on Machine Learning, 1990. doi:10.1016/B978-1-55860-141-3.50047-XMarkdown
[Lehman. "A General Method for Learning Idiosyncratic Grammars." International Conference on Machine Learning, 1990.](https://mlanthology.org/icml/1990/lehman1990icml-general/) doi:10.1016/B978-1-55860-141-3.50047-XBibTeX
@inproceedings{lehman1990icml-general,
title = {{A General Method for Learning Idiosyncratic Grammars}},
author = {Lehman, Jill Fain},
booktitle = {International Conference on Machine Learning},
year = {1990},
pages = {368-376},
doi = {10.1016/B978-1-55860-141-3.50047-X},
url = {https://mlanthology.org/icml/1990/lehman1990icml-general/}
}