【作者】 朱运磊； 岳昆； 王朝禄； 刘惟一；

【Author】 Zhu Yunlei,Yue Kun,Wang Chaolu,and Liu Weiyi (Department of Computer Science and Engineering,School of Information Science and Engineering,Yunnan University, Kunming 650091)

【机构】 云南大学信息学院计算机科学与工程系；

【摘要】 世系记录了数据产生和演化的过程,基于世系信息计算不确定性数据查询结果的不确定性,可提高处理的效率.贝叶斯网是不确定性知识表示和推理的基本框架,以有效支持基于世系的概率推理和查询处理为目标,我们利用布尔公式、逻辑蕴含和图模型的性质及等价转换机制,将不确定性数据的世系表达式等价地转换为BN,并基于BN的概率推理回答查询,设计并实现了LBNS这一基于BN的不确定性数据世系表示和查询处理系统.首先介绍了LBNS系统的结构,以及系统的关键技术,并演示了系统的主要功能和性能.LBNS给出了基于BN表示不确定性数据世系信息、以概率推理为中心的不确定性数据管理的原型.更多还原

【Abstract】 Lineages record the process of data generation and evolution.To compute the uncertainties of query results of uncertain data based on lineage can improve the efficiency of query processing. Bayesian network(BN) is the basic framework of representing and inferring uncertain knowledge.To support lineage-based probabilistic inferences and query processing,we transform lineage expressions into BN equivalently by using the properties and equivalence among Boolean formulas,logic implications and graphical models.Further,BN is used to answer queries based on probabilistic inferences.We design and implement LBNS,a BN-based system for lineage representation and inferences of uncertain data.This paper first introduces the architecture of LBNS and underlying techniques,and then demonstrates the functionalities and performance of this system.LBNS presents a prototype of representing lineages of uncertain data and probabilistic-infcrence-centered uncertain data management.更多还原