The document discusses challenges related to managing large volumes of heritage science data. It notes that while technology has increased data richness and re-use, it has also increased issues around data quality. Better quality content is now needed, including higher resolution data and richer metadata. The document explores approaches to assessing metadata quality, such as completeness and accuracy metrics. It proposes using machine learning techniques like clustering to identify topics needing enrichment based on textual features of metadata elements. The overall goal is to improve processes for quality-driven metadata enrichment.