ITR: (ASE+NHS) - (int+dmc): Constraint-based Knowledge Systems for Grids, Digital Libraries, and Persistent Archives

无论是科学收藏、教育课程储存库还是持久档案，对大量数据的管理都是科学和教育行为的重要组成部分。每种类型的数据管理系统都需要应用多种约束：数据共享（对组交互的社会学控制、访问控制）、数据显示（内容的转换迁移、组特定的视图机制）、数据联合（一致性管理）和数据控制（数据放置、数据同步）。约束可以在层次结构中组织为全局约束、基于规则的对象约束和基于过程的转换约束。本项目研究约束粒度的规范、约束语言的实现，以及支持动态约束规范和应用的数据管理系统的开发。该软件基于存储资源代理数据网格技术，该技术支持跨学术机构的数字图书馆和科学收藏联盟。该项目更广泛的影响将是开发用于分布式数据管理的自适应中间件技术。示例包括向网格环境添加服务以保持一致的元数据管理，向数字图书馆系统添加服务以验证新集合的新元数据属性，以及向保存环境添加服务以管理存档材料的真实性。我们的目标是创建通用的分布式数据管理基础设施，可以在机构层面使用，在国家规模的项目中使用，如NSF信息技术研究工作，以及在国际学术合作中使用，如全球大学网络。

The management of massive amounts of data, whether in scientific collections, education curricula repositories, or persistent archives, is an essential part of the conduct of science and education. Each type of data management system requires the application of multiple constraints: data sharing (sociological controls on group interactions, access controls), data display (transformative migrations of content, group specific view mechanisms), data federation (consistency management) and data control (data placement, data synchronization). The constraints may be organized in a hierarchy as global constraints, rule-based object constraints, and process based transformative constraints. This project investigates the specification of constraint granularity, the implementation of constraint languages, and the development of data management systems that support dynamic constraint specification and application. The software is based on the Storage Resource Broker data grid technology that supports federation of digital libraries and scientific collections across academic institutions.The broader impact of the project will be the development of adaptive middleware technologies for distributed data management. Examples include the addition of services to grid environments that preserve consistent metadata management, the addition of services to digital library systems that validate new metadata attributes for new collections, and the addition of services to preservation environments that manage authenticity of archived material. Our goal is to create generic distributed data management infrastructure that can be used at the institution level, within national scale projects such as NSF Information Technology Research efforts, and within international academic collaborations such as the Worldwide Universities Network.