Archiving Scientific Data

This report is part of the seminar Digital Information Curation held by Prof. Dr. Marc H. Scholl and Dr. André Seifert during the winter term 2005/06. Its intention is to summarize the talk held at November 22, 2005 which preceded this report. Additionally some more in-detail information about an archiving technique for hierarchical data with key structure [1] shall be provided. Section I introduces the general problem statement, shows why huge data sets are involved and why access to arbitrary versions of the data is crucial. Diff-based techniques are commonly used when it comes to data preservation with multiple versions. To get a feeling about the related work and to lay ground for the comparison between the different approaches we will explain them in section II and review the comparative test results in section IV. The main section III will deal with the archiving of hierarchical data the proposed solution published by Buneman et al. in [1] and [2].