摘要
关联规则可用于指导企业商务决策,针对关联规则挖掘的支持—置信框架会产生冗余规则的问题,该文提出了一种本体统计相关性与语义相关性相结合的关联规则挖掘方法。该方法以关联规则挖掘为目标,首先建立领域本体,并集成一个更为通用的本体系统辅助关联规则的挖掘,综合考虑本体的统计相关性和语义相关性定量计算规则相关度。应用客观兴趣度和主观兴趣度约束无趣规则的产生。与已有的方法相比,该方法有效地处理了冗余规则,实现了基于语义的知识表示。同时,该方法在心血管疾病辅助诊断系统中应用验证了其有效性和优越性。
Association rule can be used to direct enterprise business decision-making. In view of the problem of redundant rules caused by support-confidence framework, an association rule mining method and algorithm based on integrated ontology statistical correlation and semantic correlation are discussed. To mine association rules, domain ontology is constructed and integrated by a more general ontology to assist mining firstly, and correlation degree is computed by considering statistical correlation and semantic correlation. Data-driven and user-driven interest measure are used to discharge the uninteresting rules. Compared with the existing method, this method cleans out the redundant rules effectively, and realizes semantic representation of knowledge. Through the research and development of computer aided cardio-vascular disease diagnosing system, the validity and superiority of this method are validated.
出处
《南京理工大学学报》
CAS
CSCD
北大核心
2008年第4期401-405,410,共6页
Journal of Nanjing University of Science and Technology
基金
国家自然科学基金(60673060)
江苏省自然科学基金(BK2005047)
关键词
本体
关联规则挖掘
企业决策
ontology
association rule mining
enterprise decision-making
