Abstract: In order to improve the precision of Web discovery, a method named IWD(immunity-based Web discovery) for Web discovery is proposed. Mathematical formulas of self, non-self, antibody and antigen for Web discovery are established, and implementation of the method is described. The compared experimental results show that IWD has the ability of self-learning, higher precision and better time efficiency than the technique based on neural network does.

Key words: Artificial immune system, Web discovery, Search engine

摘要： 利用免疫原理，提出了一种新的Web发现方法。给出了用于知识发现的有关自体与非自体、抗体与抗原的数学描述，建立了Web发现模型IWD，并描述了该模型的实现过程，解决了传统方法无自学习能力等缺陷。对模型进行了仿真，对比实验表明IWD较传统的基于神经网络方法具有更高的准确度和更好的时间效率。

关键词: 人工免疫系统, Web发现, 搜索引擎