Newsletter Issue 5

Editorial | ECDL 2005 | DELOS Award Winner | Other Autumn Research Exchanges | European Digital Library: Catalyst for Great Change | DELOS Clusters to Contribute to a Joint Prototype | Support for Library Education in Digital Resources | TEL and DELOS Confer on Common Interests

DELOS Clusters to Contribute to a Joint Prototype

One of the central planks of the DELOS Joint Programme of Activity is global prototype development with the objective of developing a joint prototype for the future DELOS Digital Library Management System (DELOS DLMS) capable of making available results from the many groups operating within the DELOS Network. The prototype is based on existing OSIRIS/ISIS middleware further developed and extended for data streams and medical objects at UMIT with plans for further extensions at the University of Basel. The system provides generic, i.e. application- independent, services that include the registration of services and processes, interface for application development, an engine for decentralized execution of (search and maintenance) processes, service for load balancing and failure treatment.

Plans are now afoot to replace existing components and add missing functionality through realistic and relatively small-scale close collaboration among several partners in the DELOS Network which will include many short-term exchanges of scientists. This work will be led by ARCH cluster colleagues and will be based on the OSIRIS/ISIS document on their cluster website. Examples of this collaborative effort across the clusters are as follows:

• From ARCH cluster, CNR and their OpenDlib prototype, use will be made of sophisticated term extraction from text, text indexing and collection management. Padua and relevant partners will contribute from their effort on annotation services. University of Basel will add reliable sensor data management to the existing prototype.
• Work from cluster IAP, especially that pertaining in part to the CNR MILOS system will inform developments in multimedia indexing and performance evaluation when applied to different multimedia collections. Furthermore search process generation and personalization services are under development in IAP also (University of Athens) and will be incorporated as a major improvement on the existing system.
• Tasks currently undertaken by cluster A-V/NTO will serve to enhance image feature extraction, 3D shape recognition and related special indexing techniques as well as video retrieval. Work being carried forward by Vienna in respect of audio feature extraction and audio retrieval will also be incorporated.
• Cluster UIV and partners will contribute work from their tasks in respect of visualization services and visual relevance feedback which will be included as well as SOM visualization work from Vienna. Furthermore, active paper from ETH will prove another very useful service at the interface.
• Cluster KESI will devote effort on services for various transformations between standards in order to increase interoperability and the use of ontologies. For example, the Graphonto prototype system and natural language access from Crete's effort look to be very promising services in the work to improve functionality.
• Cluster PRESERV will provide preservation services.
• Multi-lingual capability will be provided by language translation services made available by cluster EVAL.
• Basel/Konstanz will take care of the OSIRIS/ISIS infrastructure that must be extended and will jointly investigate interoperability aspects as well as client services taken from the Daffodil prototype.

This list is not exhaustive. It gives examples of building blocks. Other building blocks will be identified in the course of the project and after a "call for service integration".

Initially a careful analysis of the systems architecture will be performed in respect of all building blocks that are candidates for inclusion in the DELOS DLMS prototype. The aim will be to identify, from a technical perspective, if a building block can be integrated at all, and whether this should be done in a loosely-coupled or in a tightly-coupled way. Moreover, effort required to prepare a building block for integration will need to be estimated. Following such analysis, the actual integration effort will take place in close collaboration between the organization responsible for each building block included and the task co-ordinators.

<previous   -   next>