III: Small: Rural: Querying Rich Uncertain Data in Real Time

如今，不断增长的数据量使其持续可用。示例来源包括各种传感器、RFID、计算机网络流量、电话交谈和网络搜索。这些数据中有很大一部分是有噪音的、不一致的，甚至是错误的。与确定性数据相比，不确定数据携带了更多的信息。强制不确定数据是确定性的(例如，通过接受预期)可能导致查询结果中的重大信息损失，可能导致对帮助决策的查询的错误判断。此外，基于这些数据的实时决策可能会对生活质量、经济和我们的安全产生重大影响。然而，对不确定数据的快速、实时处理是一个难题。乡村(实时查询富不确定数据)项目的总体目标是提供使该查询处理任务可行的技术。为了实现这一目标，该项目包括：(1)通过流水线将数据清洗、分布学习和查询处理集成在一起；(2)开发针对不确定数据的快速在线查询处理算法；(3)提供对不确定数据分布的丰富处理，包括压缩、共享和衡量其可靠性；(4)使用预测模型来推断不确定数据分布；(5)考虑结果的典型性来回答top-k查询。它被肯塔基州运输中心用于动态交通路由和控制应用程序，并被肯塔基大学生物医学部用于实时监控应用程序。研究成果通过肯塔基大学计算机科学研究生课程的课堂项目和会议上的教程与教育相结合。有关该项目的更多信息，请访问该项目的网页：http://protocols.netlab.uky.edu/~ge/projects/rural.

Nowadays, a burgeoning amount of data is made available continuously. Example sources include various sensors, RFID, computer network traffic, phone conversations, and web searches. Much of this data is noisy, inconsistent, or even erroneous. Compared to deterministic data, uncertain data carries more information. Forcing uncertain data to be deterministic (e.g., by taking the expectations) can cause significant information loss in query results, possibly leading to wrong judgments for queries that aid decision making. Moreover, real-time decisions based on these data can have a significant impact on the quality of life, on the economy, and on our security. However, fast and real-time processing of uncertain data is a difficult problem. The broad goal of the RURAL (querying Rich Uncertain data in ReAL time) project is to provide techniques that make this query processing task feasible. To achieve this goal the project includes: (1) integrating data cleansing, distribution learning, and query processing through pipelining; (2) developing fast and online query processing algorithms on uncertain data; (3) providing a rich treatment of uncertain data distributions including compression, sharing, and gauging their reliability; (4) using predictive models to infer uncertain data distributions; and (5) answering top-k queries with consideration of typicality of results.The RURAL project extends the state of art in real-time query processing of uncertain data for decision-making. It is used by the Kentucky Transportation Center for the application of dynamic traffic routing and control, and by the Biomedical Division at the University of Kentucky for real-time monitoring applications. The research results are integrated with education through class projects in Computer Science graduate courses at the University of Kentucky and tutorials at conferences. Further information on the project can be found on the project web page: http://protocols.netlab.uky.edu/~ge/projects/rural.