A multi-agent approach for autonomous digital preservation

Abstract: Keeping the content of digital objects accessible is a key challenge for digital archives, especially those with limited resources dedicated to preservation. This paper describes an agent-based model that simulates processes in which a digital object become obsolescent, thus a preservation action, such as the migration to the most appropriate format, is required. Agents monitor and control the local environment and deal with preservation issues individuating obsolescent formats based on global parameters such as their diffusion. They exchange information to find out the most suitable preservation action: suggestions are evaluated and propagated according to their reciprocal level of trust. The level of trust is automatically updated after every interaction through a feedback mechanism based on stigmergy. The paper shows that the framework has a stable behaviour under several use cases. Therefore this approach is suitable for digital archives that may take advantage of a multi-agent system which can either perform an autonomous preservation action or suggest a list of best candidate solutions to the user.