The Open Archive Initiative (OAI) promotes interoperability standards that aim to facilitate the efficient dissemination of content. In this paper, the authors present their ideas on sharing metadata in its original form, by applying extensible stylesheet language transformations (XSLT) on top of it, rather than converting the metadata to a specific standard. The approach proposed in this work can be applied to many scientific areas.

First, the authors present the benefits of sharing metadata via the OAI Protocol for Metadata Harvesting (OAI-PMH). These include the ability to consolidate the records from multiple providers, as well as the ability to perform incremental harvesting versus a complete harvest.

Next, the paper describes the two main modes of operation for OAI-PMH: harvester and provider. Harvester and provider agents negotiate for a specific metadata standard and exchange information in that format. The authors describe a few limitations; for example, if the provider enhances some portions of the original metadata, then it would have to replicate these changes to the mapping file that is used in the generation of the negotiated metadata standard. In turn, the harvesting agent would also have to apply a similar approach to make use of the enhanced or changed portions of the Extensible Markup Language (XML) file.

Subsequently, this work introduces an alternative approach, where the metadata is shared in the original XML format between the providers and harvester, but the XML can be accompanied by a stylesheet (XSLT). Benefits of this approach include the fact that metadata can be shared in its original format without any loss of information, as well as the fact that it becomes easy for the harvester agent to aggregate metadata from multiple sources in different standards.

Overall, this is an interesting paper with new ideas for disseminating metadata via OAI-PMH. The proposed approach seems to offer an efficient way for disseminating metadata, which is expected to reduce the duplication of efforts significantly, both at the provider end and at the harvester end. The intended audience of the paper is metadata providers as well as data aggregators.