Museum catalogues the public can actually search

A collection nobody can search is a warehouse, not an archive

The problem

Institutions with decades of collection history usually have catalogue records in a state built for registrars rather than readers: inconsistent fields, formats that have changed across different cataloguing systems over the years, and terminology that made sense to the department that entered it at the time. The public sees whatever fraction of that has been manually cleaned up and published, often a small slice of what the institution actually holds.

The usual answer

Digitisation projects sometimes tackle this by manually re-entering records into a public-facing system, which produces a clean result for whatever gets covered and leaves everything else exactly as inconsistent as before, with no clear plan for when or whether the rest gets done. A basic keyword search bolted onto the existing records returns results only as good as the inconsistent data underneath it.

How we approach it

We structure the collection data for discovery: normalising fields and terminology across different eras of cataloguing so records become comparable, and putting real search in front of the result, one that understands synonyms and related terms rather than requiring an exact match to whatever word a long-departed registrar chose. Records that cannot be cleanly normalised are flagged rather than silently dropped, so gaps in the digitised collection are visible rather than invisible.

What changes

A researcher and a visitor who is simply curious can both find the object they came for, without needing to know the exact term the original catalogue entry used. The institution's public collection becomes something people can actually search rather than a static list of what has been manually tidied up so far.