Rapid customization of an information extraction system for a surprise language
Diana Maynard, Valentin Tablan, Kalina Bontcheva, Hamish Cunningham · ACM Transactions on Asian Language Information Processing · 2003
This paper describes the rapid adaptation for surprise languages of a flexible and robust Information Extraction system based on GATE, a portable Natural Language Processing infrastructure. Our experiences show that even without a native speaker and in the absence of training data, we can quickly customize the system to a new language. We adapted the default English system for the Cebuano language in 10 days, achieving an F measure of 77.5%.