【摘要】 Web日志是一个很不完全且存在多样性特点的数据集,在获取决策规则的过程中经常会出现不一致、不完全规则的情况。提到了粗糙集理论,利用粗糙集理论在处理不完全知识上的特有优势来解决此种问题。首先把重要的用户行为特征值离散化作为属性值和值的约简,然后通过粗糙集缺省规则获取算法获得决策规则。其中条件属性的提取主要是一个对用户行为观察和分析的结果,而离散化处理方法就是应用粗糙集理论中的典型方法。这种处理方法有利于最后规则提取的进行,经过实例分析效果良好。更多还原

【Abstract】 As an incomplete and multiform data,the web logs often confront different and incomplete rules in acquisition of decision-making instance.Rough set is mentioned,utilizing rough set’extra advantage on disposal of incomplete information can resolve this problem.User behaviors’ important eigenvalue is dispersed to property value.Through reduction of property and value,then using rough set default rule acquisition arithmetic,decision-making rules is obtained.In this way,through observe and analytic behavior of web user then qualification property is distill,and the technique of disperse is one of the representative means in rough set.Because of effect is well in analysis of data,so this disperse in favor of the last process of rule acquisition.更多还原