A Context Model for Digital Preservation of Processes and its Application to a Digital Library System Rudolf Mayer, Andreas Rauber Secure Business Austria Vienna, Austria {RMayer,ARauber}@sba-research.at Gonçalo Antunes INESC-ID Lisbon, Portigal goncalo.antunes@ist.utl.pt ABSTRACT Digital preservation is an important aspect to ensure au- thenticity, traceability and auditing in processes. Digital Library Systems are one example where data transforma- tion processes are executed upon collections of data, and where such preservation of processes is an important aspect for the trustworthiness of the repository. We thus present a model for the semantic description of processes, and apply it on a Digital Library System. 1. INTRODUCTION Digital preservation deals with ensuring long-term access to digital objects over a long period, when a change in user communities or the technological environment, e.g. file for- mats, software, or hardware, would render the document otherwise inaccessible. Digital preservation increasingly broad- ens from targeting isolated digital objects, such as text or multimedia documents, towards preserving complex processes. The traceability and preservation of processes is an impor- tant aspect to verify and guarantee the authenticity of all the digital objects used and produced by the process, and to allow an analysis whether the processes were executed as expected, or according to regulations. Digital Library and Archival systems often execute processes during the lifetime of the system’s contents, from performing fixity checks on the digital objects towards more complex procedures such as preservation planning and preservation actions, such as migration of file contents to a different format. The con- tent provided in Digital Library and Archival systems may be an important part in various other processes, such as research. Thus, there is a need for making these systems auditable, and the processes they perform on their contents are of primary interest in this paper. All but the simplest processes require to be described by a multitude of informa- tion objects, as well as their connections and relations, to be successfully preserved. In this paper, we apply a seman- tic meta-model developed for describing processes and their context on a Digital Library systems. 978-1-4799-5569-5/14/$31.00 c 2014 IEEE. Figure 1: Framework for process preservation 2. A CONTEXT MODEL FOR PROCESSES A framework for digital preservation of a process is de- scribed in detail in [1], and illustrated in Figure 1. It starts with a risk assessment to identify processes worth preserv- ing, an assessment of preservation approaches, and finally a risk treatment, which is performed by digital preservation approaches. The context of digital objects consists of the en- vironment they are embedded in. It is a crucial aspect of the process preservation approach, as it provides the semantics necessary to understand the process and identify and eval- uate risks, and supports the redeployment of a process into a suitable IT infrastructure. Context can range from imme- diate and local aspects such as the software and hardware supporting the process, to aspects such as the organisation the process is executed in, service providers, and even laws and regulations. Therefore, a meta-model that allows for de- scribing these aspects, relations and dependencies, and their semantics, in a structured manner, was presented in [2]. The meta-model is designed to be adaptable to require- ments by different use cases. It is thus based on a set of smaller models that each describe semantics of specific concerns. The core ontology describes generic concepts regarding processes, and is augmented by extension on- tologies, which are describing more specific aspects. On- tologies are a well-suited method to implement this archi- tecture. The Core ontology represents a neutral, domain- independent language that is able to represent the core con- cepts of the context model. It does not address any spe- cific domain-dependent concerns, but rather cuts across the whole organisation running the process. We opted to base the core ontology on the ArchiMate 2.0 language ([3]), which includes a minimum set of concepts and relationships and the framework includes a minimum set of layers (business,