知识进化算法研究

Research on knowledge evolution algorithm

传统的进化算法是对生物自然进化过程的模拟,目前人们对它们的研究大多集中在生物自然选择层面上。将知识进化论哲学思想与生物进化理论结合起来,提出了知识进化算法的基本原理和实现途径。知识进化算法的核心思想是建立知识适应度评价函数,利用两个关键算子即传承算子和创新算子,实现知识的进化。将知识进化算法应用于图书馆读者满意度评价的知识规则进化实例中,获得了成功的结果,表明了知识进化算法的可行性和有效性。

The traditional evolutionary algorithms only simulate the course of natural creature evolution,and the current research on them mostly focuses on creature natural selection.The evolutionary epistemology idea and creature evolution are banded together in this paper,the basic principle and realization approaches of knowledge evolution algorithm are proposed.The two key operators of knowledge evolution algorithm are inheriting operator and innovation operator.Knowledge evolution is realized by the two key operators with the help of the knowledge fitness function.Knowledge evolution algorithm is used to realize the knowledge rule evolution for evaluating readers＇ satisfaction rate in libraries.The successful results show that knowledge evolution algorithm is feasible and valid.