Comparison Function for Chinese String in Entity Identification

Abstract

String comparison function is very important in the course of entity identification. Those existing comparison functions seldom considers the feature of Chinese string and the result is unsatisfied when ther are applied in Chinese string's comparison. To improve the. The experiment result indicates that our function can get higher efficient, higher recall ratio and higher precision ratio than function based on edit distance. The function we proposed can be applied to compute the similarity of strings which includes Chinese string.