Weidong Qu, Ruiqiang Zhang, Yoshinori Sagisaka
An Evaluation of Two Methods
and Their Combination for
Part-of-Speech Tagging
Abstract:In this report, we give our comparison of two models for part-of-speech (POS) tagging on
ATR treebank. One is the N-gram model and another is Maximum Entropy (ME) model. We
investigate the feasibility of ME model when using different trigger types and their contribution to
the model. We also give some results about their combination and some future work on this
research.