| A94-1007 | text styles . We have developed a | learning method | for the feature structures ( |
| A00-2017 | emphasizes the importance of the new | learning method | . Table 2 compares our method |
| C00-1048 | to make use of a data-oriented | learning method | in which linguistic knowledge |
| C00-1066 | paper , we propose an unsupervised | learning method | to overcome these difficulties |
| A97-2006 | problem can be resolved by the | learning method | without any analytical knowledge |
| A00-2015 | than the other . As a statistical | learning method | , we employ the decision list |
| C00-1046 | we propose a new unsupervised | learning method | for obtaining linguistic rules |
| A00-2017 | n-gram-like modeling . Namely , the | learning methods | make use of features which are |
| A00-2017 | types of features only if the | learning method | handles large number of possible |
| A00-2015 | , we employ the decision list | learning method | of Yarowsky ( 1994 ) . 3.1 The |
| C00-1030 | would be to use some iterative | learning method | such as Expectation Maximization |
| A00-2007 | having used the memory - based | learning method | u31-IG . ( Veenstra , 1998 ) |
| A00-2017 | machine learning and probabilistic | learning methods | used in NLP make decisions using |
| A97-1031 | subgrammars from a competence base and | learning methods | for domain-specific extraction |
| A00-2020 | experiments involving the naive Bayes | learning method | , 6213 anomalies were detected |
| A00-2015 | evidence of the decision list | learning method | as any possible pair ( F1 , F2 |
| C00-1046 | viability of our unsupervised | learning method | fi'om plain text corpora . In |
| C00-1046 | paper proposes a new unsupervised | learning method | for obtaining English part-of |
| A00-2017 | representations , along with a | learning method | capable of handling the large |
| A97-1051 | problems , our errorreduction | learning method | requires only modest amounts |
