Sorry, but a detailed description of the project is only available in german.

    Description of the Project

    In the digital age any author can put information on the World Wide Web to make it globally accessible. Scientists publish the results of their work on the web using the server of a publisher, a central preprint server or their own institutions or personal homepages. The web changes the handling of information and the use of information management systems concerning creation, distribution, retrieval and archiving of scientific results. The major problem is not publishing, but retrieval of specific information from amongst the huge amount of information on the web.

    Global search engines like Google etc. are not very specialized on a specific field. As a result, many "hits" are not relevant and the users are drowned in a flood of irrelevant answers. The field specific information a researcher looks for represents only a tiny fraction of the information available on the web.

    A solution to this problem are specialized search engines and information systems which only search in specific fields and serves only professional information.

    One examples for such a system is PhysNet, the physics department and document network ( PhysNet is a so called distributed information service and one of the largest and completest portals to the learned field of physics.

    There are many requirements, which information services with distributed or decentralized sources, like PhysNet is, have to comply with.

    In such systems one find very heterogenous sets of data and document sources and the handling with this must be well-regulated. Different kind of information and document-types have to make accessible and retrievable. In this connection, the networking of different fieldspecific information-services (also the interdisciplinary networking) is a very up-to-date topic. For this, a standardized vocabulary of Metadata (e.g. DublinCore) for the coverage of relevant (and professional) information in needed as well as standardized protocols for the exchange of data between different services. Another requirement is, of course, that queries to such an informationsystem have always to be processed and answered quickly, independently from any problems with the network load or even the failure of Software or Hardware.

    To build such a system is the aim of the project SINN - Searchengine-Network in the International Natural Science Network. (the german word Sinn means sense).

    In more detail: SINN will enhance the distributed information system PhysNet to a fast and secure service by setting up a network of independent but cooperating Harvest-brokers and Harvest-gatherers as a searchengine system and building a network of PhysNet-mirrors.

    SINN is supported by the German Academic Network Organisation (Deutsches Forschungsnetz e.V. DFN), with the financial support of the German Ministry of Education and Research (Bundesministerium für Bildung und Forschung - BMBF) and the Government of Lower Saxony.

    This project is run by the Institute for Science Networking in Oldenburg in cooperation with the international partners of the PhysNet-Service which agree to set up mirrors.

    The idea is to set up an active network of PhysNet-mirrors and of a Harvest-based searchengine system. For this the development of a language or protocol (based on XML-Query) for data exchange and the implementation of a communication between the distributed search-engines is necessary.

    The integration and training of the international partners is an important point in this activities as well as a close view on international activities of different initiatives like the WorldWideWeb-Consortium (W3C) - or its workinggroup XML-Query- , the Open Archives initiative (OAi) and others and collaborate with them.