Digital Collections Risk Assessment at LSE: Using DRAMBORA

At LSE we are undertaking a programme of infrastructure development to build on our capacity to store, manage, preserve and provide access to our growing digital collections. We have a mature IR service holding our research outputs, but are seeking to expand our capacity to handle other digital collections such as the outputs from digitisation projects and born-digital archives.

Key to this development is making the case for investment, a task which is made even harder by the current economic outlook in higher education. DRAMBORA, a tool introduced on the KeepIt course in which LSE participated, takes a risk assessment approach to auditing repository contents (which can be taken to mean any collection of digital material) which highlights areas where a repository needs to develop its practices. “Practices” can mean developments to technology, skills or organisational attitude.

The tool itself is both a paper-based exercise and an online assessment centre where you can go through the process of self-auditing digital collections. The process has already been described in other posts on this blog. One point to note is that the tool advises that you will need access to various supporting information, such as institutional strategies, policy documents, mandates and staffing structures. However, once this domain modelling is in place, the core of the assessment is analysing the risks to your collections, the process and outputs of which are not tremendously enhanced by the use of this information.

DRAMBORA is based on various tools such as OAIS and TRAC (see the full list) and comes with a set of predefined risks which have been developed and refined by the project partners, and which can provide the starting point for adaptation to local circumstances. In our case, we took a fairly lightweight approach and worked through the provided list to select a small number of risks that crossed functional domains (organisational, technical, and so on) and were representative of concerns in each area. This included high-level, intangible risks such as threats to institutional reputation should we fail to preserve our digital collections, through to low-level, technical risks such as failing to preserve a significant characteristic of a file format or make suitable backups. Our purpose was to demonstrate the general state of our collections and to highlight areas where we need to focus our efforts, rather than to capture every detail in all or even one functional area. It may be that we will use the tool in more detail further down the line, producing, for instance, a comprehensive list identifying technical risks.

We ended up with a set of about 10 risks, which is by no means exhaustive, which we refined with local expressions of each risk by taking examples from our own circumstances and collections. This gave us a good oversight of our current situation and also provided the basis on which to sketch out a roadmap for development. We found the ability to scale our use of the tool to this fairly lightweight implementation to be a huge benefit of the bottom-up or self-audit approach.

The outputs, in the form of a risk register, gave us a succinct, well-reasoned and clearly explained summary of the risks to our collections, given an air of authority firstly by the nature of the process—a thorough and transparent audit—and secondly by the provenance of the tool—its development by a group of experts in digital preservation and association with a range of best practice in the field.

We took the risk register produced by the tool and combined it with parallel work of the infrastructure development programme which was investigating our functional requirements and possible software solutions to support our curatorial activities. This resulted in a report and set of recommendations for our senior management team which identified the risks our collections face right now, and proposed solutions. In each case we were able to demonstrate that our proposed developments addressed specific risks to our collections, users and institution generally.

DRAMBORA formed part of our investigations of our next steps in building preservation capacity, and was invaluable in the role it provided—making the case for investment clear by pointing out the risks to our current position. However, our use of the tool formed part of an ongoing programme of work by an established team which included other strands of exploration, and the importance of this wider context cannot be underestimated. Nevertheless, the self-auditing or bottom-up approach to understanding necessary developments in terms of risk has proved very beneficial.