OUTLOOK: NETWORKED ARCHIVES

Besides consolidating netzspannung.org as a training and skilling instrument, future development also aims to network the platform with the archives of other media-cultural institutions. In this way the knowledge gathered in the various archives of media art should become researchable online at a single point, accessible via one common interface. Then users will no longer have to search through all the websites of the different archives, constantly keying in the same search request. Instead they will be able to input their search request via the interface of the networked archives and obtain the search results from all the combined archives in one overall view.

In order to exchange data within such a network, a technical link will have to be established between the archives. In addition, it will be necessary to adapt the data of the different archives according to an overall information structure.

Connected Archives

To create a data pool across platforms and the network, a common communication interface will be introduced, enabling data to be uniformly accessed and exchanged among different implementations. To this end, netzspannung.org is developing an archive framework on the basis of web services, which will permit remote Internet archives to be integrated, and which can be expanded through specific plug-ins for each archive. The basic task of the plug-in is to transform the function calls for each implementation and to map the data formats. The result will be a network of archives, in which data can be transparently retrieved and stored.

Dynamic Adaptation

In order to interrelate the content of the various archives in a semantically correct manner, their different structures will have to be appropriately mapped with one another. However, given that the operators of the various archives cannot realistically be expected to agree on a common content system, a system is being developed under the auspices of netzspannung.org which will recognize and link semantically related attributes and structures (e.g. details of authors, titles etc.).

In the semantic web research area, so-called metadata are being used to do this, which more precisely describe the actual user data. However, because these are also differently labelled in the archives, a data adapter is undergoing development, which will help the operators of the archives to match up the structures and semantics of different archives. The concept envisages, among other things, using thesauruses and analyzing whether structural similarities exist.

Moreover, with each networking the system will evaluate the accumulating metadata in order to advance automatization and help integrate further archives. In this way mapping rules will be created, which in future will permit archive data to be dynamically mapped to the form required by the inquiry interface.