【摘要】 根据概率数据的描述形式对概率数据分为基于关系的概率数据模型和基于XML的概率数据模型两类。基于关系的概率数据模型是为每个元组引入概率标记属性表示不确定性,使元组的存储、查询处理变得复杂;基于XML的概率数据模型是在普通XML树中添加表示概率属性结点,能够表示多粒度的概率信息。设计了映射为关系的概率XML数据的与PDTD无关的PXRel和PXParent两种存储模式,并通过实验验证了其有效性。更多还原

【Abstract】 According to the embodiment forms of probabilistic data,probabilistic data is classified two kinds of probabilistic data model what are based on relation and based on XML.The method of the probabilistic data model based on relation is to introduce probabilistic label attribute to every tuple to express the uncertain information,so the storage and query process of the tuples becomes more complexity than before.And probabilistic data model based on XML which is appended the probabilistic attribute nodes in ordinary XM tree can naturally express probabilistic information of many granularities.Two storage modes mapping the XML documents to relation model are designed what are PXRel and PXParent independent on PDTD. The experiment results show the validity of the storage ways.更多还原