Congruence Engine: The Summation
2025-04-10_The_Summation_CEBlog_FINAL.pdf
2.18 MB
Adobe PDF
gF/ZSlBvYzI=
We want to make all UK heritage collections fully accessible together, not just institution-by institution as now, to those who would use them. The means to achieve this are their digital records. But, because of the history of heritage organisations, our current records are ill-suited for this purpose: in the main, they are thin, inconsistent and isolate collections items from the contexts in which they make sense. In Congruence Engine, we have therefore experimentally created ways to link collections with historical data as well as with each other, techniques that in various ways overcome these weaknesses. Because the creation of useful collections linkage requires the agency of its potential users, and because computing techniques are fallible, UK collections linkage requires a social machine for the task: a human-technical approach that mobilises human motivation and thereby overcomes machine incapacities. Many of our techniques employ machine learning, with the result that we have within our grasp the capacity to overcome decades of limited investment in collection documentation. But, because we are using these techniques, we confront a series of ethical issues that demand to be researched, even as we press for the resources to create user interfaces and scalable prototypes for a future infrastructure, which together can reveal the potential of what we have glimpsed.
| Funder name | ISNI | Awards |
Arts and Humanities Research Council | ||
