G22.2591 - Advanced Natural Language Processing - Spring 2004
Name Recognition, cont'd
Unsupervised learning of names
The methods described to date require the preparation of a substantial
amount of training data. Can this requirement be reduced?
We will look today at several papers which investigate this issue:
All of these efforts aim to produce taggers with moderate performance
starting with minimal resources. In principle, it should be
possible to use similar techniques to improve the performance of good
taggers (taggers with substantial training data) or to adapt good
taggers to shifts in domain, but (as far as I know) these crucial
issues have not yet been explored.