Distributed data mining services leveraging WSRF

The continuous increase of data volumes available from many sources raises new challenges for their effective understanding. Knowledge discovery in large data repositories involves processes and activities that are computationally intensive, collaborative, and distributed in nature. The Grid is a profitable infrastructure that can be effectively exploited for handling distributed data mining and knowledge discovery. To achieve this goal, advanced software tools and services are needed to support the development of KDD applications. The Knowledge Grid is a high-level framework providing Grid-based knowledge discovery tools and services. Such services allow users to create and manage complex knowledge discovery applications that integrate data sources and data mining tools provided as distributed services on the Grid. All of these services are currently being re-designed and re-implemented as WSRF-compliant Grid Services. This paper highlights design aspects and implementation choices involved in such a process.

The continuous increase of data volumes available from many sources raises new challenges for their effective understanding. Knowledge discovery in large data repositories involves processes and activities that are computationally intensive, collaborative, and distributed in nature. The Grid is a profitable infrastructure that can be effectively exploited for handling distributed data mining and knowledge discovery. To achieve this goal, advanced software tools and services are needed to support the development of KDD applications. The Knowledge Grid is a high-level framework providing Grid-based knowledge discovery tools and services. Such services allow users to create and manage complex knowledge discovery applications that integrate data sources and data mining tools provided as distributed services on the Grid. All of these services are currently being re-designed and re-implemented as WSRF-compliant Grid Services. This paper highlights design aspects and implementation choices involved in such a process. (C) 2006 Elsevier B.V. All rights reserved.