Research on Parsing Technology in Machine Translation System
Yao Meng · Journal of Heilongjiang Institute of Technology · 2002
Parsing is a central problem for many tasks in natural language processing,especially for machine translation.This paper presents the experiences in parsing for several machine translation systems,and proposes a modular parser using combination of statistic and rule methods.It integrates several strategies including Hide Markov method,statistic decision tree,history-based methods and employs semantic information in the algorithm.Experimental re-sults indicate that this approach is successful in both English and Chinese parsing.