钟对基于Web页面信息本体的信息抽取中,需人工根据待抽信息项的概念和对应的实例值来建立本体的缺点,设计一个页面信息本体的自动学习方法.论文利用前期在基于DOM的页面相似路径归纳学习算法和基于PAT-tree的自动关键词识别算法上的研究成果,使用改进的TF·IDF统计方法和复合事件的关联规则算法完成概念和概念间关系的学习,建立页面信息本体,减少建立本体的人工工作量.%According to the weakness of building ontology needs manually designated concepts and instances from the basic information of web, an automatic ontology learning approach based on web information items is designed. Utilizing preresearch that an arithmetic of an inductive learning based on DOM for the similar path of information items and an approach for identifying automatic keyword based on PAT-tree, the learning for the concepts and the relation between concepts is implemented by using an approved TFIDF statistic method and an algorithm of composite event association rule, the information item ontology is built, the manual workload for building ontology is reduce
展开▼
