中文摘要：

为发现Web使用记录中所蕴涵的用户访问模式，在深入分析日志本体中事件间的抽象关系后，提出适用于原子事件和复合事件间整分关系推理的ALC传播规则扩展已有的推理模式，并在此基础上提出一种挖掘日志本体的ILP方法。该方法结合描述逻辑和Horn规则在知识表示和推理过程中互补的特点，采用AL-log混合系统构建知识库，利用约束SLD－反驳消解和扩展ALC传播规则从日志本体中学习用户访问模式，达到站点商业智能和个性化的目的。最后给出验证该方法的实例，实验结果表明了该方法的可行性和有效性。

英文摘要：

In order to discover the user-access patterns in the Web usage records, this paper analyzed the abstract relations between events of log ontology and extended ALC propagation rules for reasoning the part-whole relation between atom events and complex events. Based on the new rules, this paper proposed an ILP approach to mine the log ontology. This approach which integrated description logics and Horn clause rules, adopted AL-log hybrid representation and reasoning system to build the knowledge base, and used constrained SLD-reputation and extended ALC propagation for learning user-access pattern from log ontology. It made business Web sites more intelligence and personalization. The experimental results show that this method can help site owners to create access rules effectively and it is quite feasible to solve practical problems.