基于遗传算法的本体概念分类规则学习算法

为了更好地界定本体中的概念，提出一种基于遗传算法（Genetic Algorithm，GA）的本体概念分类规则的学习方法．从已有的本体库中获取实例作为训练样本，通过该算法寻找一组与数据样本集一致的规则．以一组规则集作为遗传算法的个体，即优化的目标，同时考虑到规则集的覆盖性、一致性、简洁性和多样性4个方面建立适应值函数，优化得到一组能够分类概念的规则集合．进而这组规则集可用于指导和丰富本体知识，例如当本体中引入新的实例时，可以通过此概念分类规则集确定实例所属的概念．对已有本体学习后的实验结果表明该算法收敛性很好，而且能获得较好的规则集．

Ontology concept classification rule learning algorithm based on genetic algorithm

To determine the concepts in ontology more accurately, an ontology concept classification rule learning method based on genetic algorithm is presented. The instances are from existing ontology as training samples, and a set of rules that are consistent with the sample sets is found out by the algorithm. A set of rules are optimized as an individual of a genetic algorithm, and the fitness function is defined considering the rule set's coverage, consistency, simplification and diversity. And a set of optimized rules of concept classification are obtained, which are used to direct ontology enriching. The Experimental results with an existing ontology show that the algorithm has good convergence and can obtain optimal rule set.