带有不确定性的数据集成

DATA INTEGRATION WITH UNCERTAINTY

传统的数据管理技术没有考虑现实中存在的不确定性，很多现有的方法选择最可能符合用户意图的结果，从而消除了不确定性问题，但有时这些不确定信息是有价值的，忽略掉这些信息可能使对用户有用的信息不能返回给用户．因此，笔者阐述了产生不确定性的三个方面的原因，并提出了一个使用用概率数据库和概率映射方法处理不确定性的数据集成方法．

The traditional data and many of them select the results management technology didn＇ t consider the uncertainties existed in reality, satiring the users＇ intention to get rid of the uncertainty. But sometimes this uncertainy information is of great value, usefull information may not return to users if ignoring them. Therefore, this paper illurstrates three reasons of uncertainty, and proposes an approach that dealing with uncertainty in data integration by using probablity database and probablity schema mapping.