Munich Working Group on Digital Long Term Preservation
Meeting on October 25th, 2007
Summary
The Bavarian State Library, Munich, and the Universität der
Bundeswehr München organized this meeting with the focus on
migration of digital objects. The Institute for Software Technology was
pleased to welcome the members and to present three lectures. The first talk sketched the core problems of long term preservation and illustrated the three major strategies for preserving digital information, i.e., hardware and software museum, emulation, and migration. Pros and Cons of migration were briefly discussed. In order to get a deeper and clearer understanding of migration, a more formal framework was introduced. It is based on the ideas described in OAIS - a reference model for an open archival information system. The interpretation of bit streams (data) by so called representation information in OAIS's information model is detailed by a layered model which mainly consists of representation layers and associated mappings. These mappings can either be realized by operations (mainly software procedures) or documented by rules in abstract ways. Based on the different layers of representation different types of migratation can be isolated and studied. Since there are very critical types of migration, strict and formal approaches are required for trustworthiness and economic viability.
The second talk demonstrated such a formal plug-in for our framework. In order to make migration verifiable, significant properties of digital objects and preservation requirements related to these properties have to be expressed in a formal language. Significant properties are concepts that are vital for the preservation of information. Therefore, any implementation - before and after the migration - has to represent these concepts. Means for the matching of concepts and implementations and for the automated verification of preservation requirements are the core features. Formal logics and automata theory yield the foundations. A prototypical tool connects implemented migration routines to the formal framework. Recently, the approach has been validated for structural website transformations. Collections of XHTML documents were transformed to ODF (Open Document Format, an ISO standard) and vice versa while preserving content, CSS-based layout (Cascading Style Sheets) and link-consistency. It turned out that formalization is indeed helpful here and supports in setting up trustworthy migration routines.
Finally, the third lecture presented the current architecture of mediaTUM, a repository system designed for the needs of a large university. Requirements are usability, flexible workflows for the ingest process, customizable schemes for object collections and metadata, long term availability of information, as well as the capability to integrate the repository into a broader application architecture. This open software project is coordinated and realized by the Technische Universität München (Technical University of Munich - TUM).
The Institute of Software Technology thanks for the numerous contributions, suggestions, and the patience for listening to some details of computer science.
The Munich Working Group on Digital Long Term Preservation offers an efficient platform for discussions and knowledge transfer in the domain of digital long term preservation by benefiting from the exceptionally high concentration of libraries, archives, museums, universities, scientific societies, computing centers, political foundations, chambers of commerce, broadcasting houses, and high-tech companies in the area of Munich.
