A quarter of online journal articles are not digitally archived, warns Martin Eve
Principal R&D developer at Crossref, Martin Eve’s research reveals a significant challenge in digital preservation within academic journals. His analysis of nearly 7.5 million journal articles found that while 58% are archived, almost 28% have no digital preservation, putting a large portion of scholarly work at risk of disappearing. This issue stems partly from a shift in responsibility from libraries to publishers, many of whom may not fully realize their role in preservation. The study suggests more education and resources are needed to safeguard the scholarly record. Full article published on Research Professional News.
This is no surprise to me. There is a serious lack of useful documentation on how to set up archiving. Moreover, unless you are running OJS the only option is to join an organization like CLOCKSS or Portico. Many online journals run on a shoestring budget and the cost of joining a preservation network is nontrivial. For example, joining portico costs $250/year minimum, whereas our journal runs on a machine for which offsite backups cost $2.70/month on digitalocean. Theoretically we could install and run LOCKSS, but they recommend at least 8 cores and 16GB of RAM. By contrast, our journal runs on a machine with 1 core and 2GB of RAM and serves the traffic just fine at 6% utilization. These preservation systems are not what you would call lightweight or well-engineered. Digital preservation is apparently not interesting enough to attract serious engineering work.