Not all relationships within a collection are directly visible in metadata or descriptions. Objects can be connected through shared motifs, themes, forms, materials, historical context or symbolic meaning. These associative relationships are valuable for research, presentation and public storytelling. With Associations, we help make such relationships visible by combining visual characteristics and contextual information.

We help detect thematic and associative relationships between objects based on visual and contextual meaning. We look not only at descriptive metadata, but also at similarities in image, subject, material, use, period or cultural context. This reveals relationships that might otherwise remain hidden and creates new opportunities for collection research, narrative access and exhibitions in which objects are connected on content-related or visual grounds.

Control and quality

AI does the preparatory work. Results remain reviewable and are configured around your sources, fields, terminology and quality criteria. Where relevant, the source or passage remains traceable. Your organisation decides what is adjusted, approved and ultimately used.

A concrete first step

We prefer to start with a clearly defined test set. Together we agree the required output, quality criteria and review flow. This shows what works before you scale up.