Characterising AI Models for Cataloguing

This paper examines how AI models can support cataloguing for digital collections, a task that follows digitisation and often depends on slow, costly expert labour. The authors compare different implementations and models, then provide both qualitative and quantitative evaluations of their experiments. The paper also develops recommendations intended to extend beyond this particular cataloguing use case. The supplied abstract does not specify the evaluated models, datasets, metrics, results, or recommendation details.