Abstract: Existing Artificial Immune System(AIS) for recognition applications in the text, the detector generating algorithm for different genes, such as the quality of treatment, there can not be the optimal response gene frequency in the antibody deficiency. This paper proposes the concept of degree of gene dominant through the detector generation algorithm and the matching algorithm introduces genes dominant degree factors to improve the efficiency of the algorithm. Experimental results show that the introduction of dominant degree of the detector generating algorithm can be reduced by 30% of the time complexity.

Key words: Artificial Immune System(AIS), text recognition, matching algorithm, detector generation, dominant degree

摘要： 现有的人工免疫系统被应用于文本识别中时，检测器生成算法对不同基因等质化对待，不能最优反应基因在抗体中出现的频率。针对该问题，提出基因显性度的概念，通过在检测器生成算法及匹配算法中引入基因显性度的因子来提高算法效率。实验结果表明，显性度的引入可降低检测器生成算法约30%的时间复杂度。

关键词: 人工免疫系统, 文本识别, 匹配算法, 检测器生成, 显性度

CLC Number: 

• TP18