Maintaining object authenticity in very large digital libraries

Abstract

Digital libraries are increasingly being used to manage re- search data, leading to a signi¯cant increase in the amount of material held in such systems. Much of this material is irreplaceable, and there is a pressing need to maintain long-term access to it; automation of cura- tion is key if a scalable solution is to be found. The concept of signi¯cant properties is key to maintaining the integrity and authenticity of a dig- ital object over time and technological change. We present a case study which addresses signi¯cant properties as they apply to a working scien- ti¯c community active in experimental and computational research, and we describe an approach to automating this model by representing cura- tion policies as rules that are implemented using the iRODS middleware.