BEGIN:VCALENDAR
PRODID:-//hacksw/handcal//NONSGML v1.0//EN
METHOD:PUBLISH
BEGIN:VEVENT
DESCRIPTION:Click for Latest Location Information: http://edw2010.wilshireconferences.com/sessionPop.cfm?confid=38&proposalid=2355\nWe are proposing a new approach to data integration that we call Data Description Framework (DDF)[1-4]. In its core is a unified abstract data model that imposes no restrictions on the source-subject of integration: neither on its model, nor data types, nor data modality. Original data and data-semantics can be accommodated within the DDF without loss and or distortion (unless chosen to perform data staging) typical for traditional data integration approaches. The approach offers important benefits in the environments where diverse sources need to be integrated in an ad-hoc fashion, e.g. military applications. Such integration and the resulting Data Space require new approaches to Data Management (DM) - approaches that are as dynamic and flexible as the integration approach itself. \nAmong the vast number of areas that are subject to DM we concentrate on Metadata Management (MM) that enforces Data Quality, Reference and Master Data, and Data Movement (Extraction, Transformation, Load). Because the resulting Data Space does not adhere to a specific domain data-model and is supposed to accommodate data from various, often not known a priori, domains, we need new MM approaches and policies:<ul>\n<li>Post-factum (or rather post-Load) and continuous Data Quality control. Generally, at the moment of Load, the Data Quality requirements for the loaded sources are not known, and it is impossible to control data quality. Data quality control continues as the new data sources are integrated into the Data Space and the Reference and Master Data evolves.</li>\n<li>Evolving Reference and Master Data that are subject to integration itself. To handle the scale and diversity, it is important to define and implement integration and evolutionary patterns of Reference and Master Data.</li> \n<li>Meta Metadata Management. Tools, rules and approaches that define the evolution mechanisms and patterns of MM to regulate MM and keep it from becoming unruly and messy in the heterogeneous and dynamic environment.</li></ul>\nThe new MM approaches are being implemented within a DoD intelligence data integration project.
DTSTART:20100317T080000
SUMMARY:Rethinking Data Management in an Evolving Integrated Data Space
DTEND:20100317T082959
LOCATION: See Description
END:VEVENT
END:VCALENDAR