Digital Preservation Processes: Archival Judgement and Automation

(By Ruud Yap, National Archives of the Netherlands)

So many presentations, so little time. What sessions does one choose. I guess one should go for the sessions with practical solutions to everyday problems. The excellent session of Robert Sharpe “Digital Preservation Processes: Archival Judgement and automation” proved just to be that. The session concerned the issue of preservation, the potential huge amounts of electronic records and the possibilities for automation. To be precise Sharpe described a preservation process and the opportunities for automating this process.

His presentation brakes down quite logically by stating that a automated system is as smart as the person who gives input to that system. The description of the preservation process is visualised by the PLANETS framework:

The first step is gathering all your requirements and criteria for the preservation of records. Secondly you identify what records you keep. Then you plan the actions you should perform and you act. This all sounds pretty straightforward and the good news is: most of these steps can be automated!

But the archivist isn’t let off that easily. Automation is only possible after intense and fundamental human interaction. To put it bluntly: without an archivist there is no automated preservation.

Mister Sharpe clearly described the human involvement that is needed to perform the digital preservation process. The involvement of the archivist can be broken down to the decision what criteria are vital towards the preservation of records.

The archivist defines what it means for a record to be obsolete.

The archivist decides what to do about obsolescence.

The archivist describes what factors define a successful preservation action.

These criteria have to be made machine-readable and then, and only then preservation can be made automated. The major difficulty however is the quantification of opinions. Some criteria are based on opinion and not necessarily fact. Or to put it differently: some decisions and criteria are not necessarily objective.

This problem is recognised by Sharpe and as a solution he shows that such criteria are quantified and also provided with scaling. This way tolerances can be built in the processes. Opinions are quantified, objectified and documented.

Robert Sharpe presented us with a detailed session on the achievements made in the PLANETS project in which he showed us the possibilities of automated preservation of digital records. A good framework exists within PLANETS which will grow wiser and more robust over time. He also showed us that although much can be automated the human –archivist- factor can never be ignored.