Abstract
Several scientific communities relying on e-science infrastructures are in need of persistent identifiers for data and contextual information. In this article, we present a framework for persistent identification that fundamentally supports context information. It is installed as a number of low-level requirements and abstract data type descriptions, flexible enough to envelope context information while remaining compatible with existing definitions and infrastructures. The abstract data type definitions we draw from the requirements and exemplary use cases can act as an evaluation tool for existing implementations or as a blueprint for future persistent identification infrastructures. A prototypic implementation based on the Handle System is briefly introduced. We also lay the groundwork for establishing a graph of persistent entities that can act as a base layer for more sophisticated information schemas to preserve context information.