Evaluation of Preserved Scientific Processes Rudolf Mayer 1 , Mark Guttenbrunner 1 , and Andreas Rauber 1,2 1 Secure Buisness Austria, Vienna, Austria 2 Vienna University of Technology, Austria Abstract. Digital preservation research has seen an increased focus is on objects that are non-deterministic but depend on external events like user input or data from external sources. Among those is the preserva- tion of scientific processes, aiming at reuse of research outputs. Ensuring that the preserved object is equivalent to the original is a key concern, and is traditionally measured by comparing significant properties of the objects. We adapt a framework for comparing emulated versions of a digital object to measure equivalence also in processes. 1 Introduction Digital data is in its nature volatile, and always needs an environment in which it is rendered to a form that makes it useful. The interpretation of the bitstream is the subject of digital preservation (DP) research. Two strategies have evolved as the most promising. Migrating a digital object means continuously changing its format to one that is not obsolete at the time of use. Emulation keeps the original digital object but changes the rendering environment, by modifying the applica- tion used to render the object, or replacing the original hardware by introducing a virtual layer, using the original software-stack for rendering. It is necessary to evaluate that the result of the digital preservation action produces a rendering that is similar to the original in its “significant properties”, i.e., the properties of the object deemed important for future use by the designated community. For migration, the significant properties of the object extracted from the original and migrated form are compared. With emulation, one has to compare the ren- dering of the digital object in the two environments. Traditionally DP research concentrated on objects that behave deterministic, i.e., are rendered similarly on the same system during each rendering, such as text-documents, videos, im- ages, or database content. Objects that are rendered non-deterministic change their rendering depending on user input, hardware values, or random values, e.g. digital art or computer games. The preservation and curation of scientific data is important to ensure reuse and long-term usability of data that has been the basis to scientific experiments. Data, however, does also require information about its context and the processes involved in its creation and usage, e.g., the setting where it was created or interpreted. It may be impossible to recreate the original experiment, and thus a preserved process that allows to reproduce and trace how results and analysis were obtained is important. eScience processes are depending on specific software or hardware, thus facing digital obsolescence. To