This first blog post will show you a model that originates in the design of the system iRODS, but can be adapted into a broader scope of defining the relationships in data management. The model is posted in the paper “Concepts in Distributed Data Management or History of the DICE Group” as figure 8 on page 15.
So with one’s arm twisted the model can be used to find the relation between purpose, collection, policy etc. The model is created for being able to make computer based ruling, but seems to make sense in a broader perspective.
What is nice about the model is that it assumes that something has a purpose. In many other generic model the purpose is a missing – but nevertheless – important point of view. The collection of data (which is made up of digital objects) is through various properties bound to a policy. This gives the ability to differentiate between e.g. dark archives and public available repositories.
Policy
On a policy level, you can use this model to discuss the policies that should apply to your collections, keeping the purpose in mind. Once you have the purpose of why you are making collections, you can define the properties, which then again affects the policy. Based on this work you can also start to talk about which procedures that should be present to enforce the policy.
Infrastructure
The properties and procedures which are both related to the policy, can be used as guidelines for the needed infrastructure.
Competence
This model does not give any guidelines on levels of competence.