Abstract: There is the fault of blindness in traditionary Chinese word sense disambiguation algorithm based on primitive
co-occurrence data.This thesis puts forward a“Double-Distance”word sense disambiguation algorithm，which considering two
parameters of distance when calculates these relation-modulus between the maltivocal word and the character-words，the
space distance between the character-words and the maltivocal word，the space distance between the currently maltivocal
word and the same maltivocal word which has been selected sense at the latest.The experiment shows that the modified algorithm
are effective.

摘要：

针对传统基于义原同现频率的汉语词义排歧算法的“盲目性”，提出一种“双距离”词义排歧算法，即在计算待排歧词各义项与特征词之间的相关系数时，考虑两个距离因素：特征词与待排歧词之间的空间距离；最近选择该义项的同形歧词与该待排歧词之间的空间距离。实验表明，改进的算法是有效的。