Abstract: E-commerce website evaluation depends on the knowledge rules to a large extent. In this study, the evaluation index system of E-commerce website is established and the representation method of knowledge rule is analyzed firstly. Then, a method for mining knowledge rules of E-commerce website evaluation based on an improved genetic algorithm is proposed. In the algorithm, selection operator, help operator, crossover operator and mutation operator are used to generate new knowledge rules. Knowledge rules are evaluated by their accuracy, coverage and reliability. Experimental results show that this method for mining knowledge rules is feasible and valid.

Key words: knowledge rule mining, knowledge rule base, fitness of knowledge rule, E-commerce Website evaluation, genetic algorithm

摘要： 电子商务网站评价知识规则是对电子商务网站的运行情况和工作质量进行评价的重要依据，优质、合理的知识规则将使评价更加公正、更加客观。在分析并建立电子商务网站评价指标体系的基础上，将一种改进的遗传算法用于电子商务网站评价的知识规则挖掘，提出了一种基于遗传算法的电子商务网站评价知识规则挖掘方法。该方法利用选择算子、助长算子、交叉算子和变异算子来产生新的知识规则，使用正确度、覆盖度和可信度来对知识规则进行评价。实例表明，这种方法在进行知识规则挖掘时是完全可行的和有效的。

关键词: 知识规则挖掘, 知识规则库, 知识规则适应度, 电子商务网站评价, 遗传算法