EAGER: Collaborative Research: Conflict Resolution and Exchange of Temporal Data

点击翻译按钮获取中文摘要

An increasing number of applications, such as fraud detection and personal electronic health records, all rely on the availability of an integrated, consistent, and query-able view of entities over time. Such a view, implemented as an integrated archive, consolidates temporal information from different data sources and creates significant added value by making explicit the times when facts about the entities are true. The task of creating an integrated archive remains non-trivial even with the state-of-the-art data integration techniques. This is a high-risk research project that targets challenging problems in the nexus of several research areas: data exchange, temporal databases, conflict resolution and data inconsistency. This project has the potential to make significant societal impact through the development of a foundational framework that facilitates the process of consolidating temporal data from different heterogeneous temporal data sources. It is anticipated that the availability of such a platform will lead to new classes of applications, particularly those involving profiling, tracking, monitoring, and understanding of the evolution of entities over time. This project will involve the research and training of graduate and undergraduate students in science and engineering, including members of groups underrepresented in computer science, in pertinent areas of data integration. Results obtained from this investigation will be incorporated into advanced graduate data management courses on large-scale data integration that will be developed and taught at both SUNY at Buffalo and UC Santa Cruz. Selected material from the outcomes of this project will also be incorporated into upper-division undergraduate database courses taught at both universities. This proposal will develop formal foundations and principled solutions to the problems of exchanging temporal data (i.e., transformations of temporal data from one format into another) and resolving conflicts of temporal data. The unique challenge in managing temporal data that underlies all aspects of this proposal is the management of the discrepancy between the abstract declarative view of temporal data and its corresponding concrete implementation view. The general approach adopted by this proposal will be two-tiered: data exchange and conflict resolution over temporal data are declaratively specified using the declarative point-based view, while computational tasks required to exchange data and resolve conflicts manipulate the underlying time intervals. Results from this investigation will also lay the groundwork for longer-term research such as answering queries over inconsistent temporal databases. For further information see the project web site at http://greenwich.cs.ucsc.edu or http://www.cse.buffalo.edu/~chomicki/.