【摘要】 给出了一种将基于统计与基于词典方法融合而成的混合中文分词方法,利用统计方法进行词典的动态扩展,克服了基于词典算法对"完全词典"的依赖;利用词典对统计的一些参数进行估计,避免了以往靠实验得到这些参数的不确定性,同时将 RMM 算法与统计算法结合起来.最后通过仿真说明该混合算法比传统的基于词典和基于统计的方法更有效.更多还原

【Abstract】 A hybrid approach to Chinese word segmentation which combines two kinds of word segmentation algorithms based on statistics and dictionary together is presented.The dic- tionary used in segmentation is extended dynamically using the approach of statistics which conquers the dependency on the"complete"dictionary,and the dictionary gives an appropriate estimation for some parameters which is important in statistical algorithm such that the uncer- tainty brought by the experiments is eliminated.Moreover,the RMM and the statistical algo- rithm are combined together.At last,the simulation results show that the proposed hybrid ap- proach outperforms the one which is purely based on dictionary or on statistics.更多还原