Requirements modelling and evaluation for digital preservation

摘要

Most methods for the general problem of Commercial-off-the-shelf component selection use goal-oriented requirements modelling and multi-criteria decision making techniques and are applicable across a wide range of domains. This usually implies high levels of complexity. Recently a very specific selection problem emerged in the context of digital preservation. The selection of the most suitable tool to keep a type of digital object alive when the original technical environment ceases to exist is a highly complex domain-specific selection problem with several peculiarities: Highly homogeneous functionality across tools, complex evaluation of quality across settings, and a high need for automation, standardisation, and documentation. This paper describes an evidence-based empirical methodology for COTS component selection in digital preservation through controlled experimentation. We describe the specific selection problem, show how the process of utility analysis can be tailored to fit the problem space and describe the methodology, which is geared towards automated evaluation in an empirical setting. We outline existing tool support and discuss case studies and future directions.