video1.0<iframe src="https://www.loom.com/embed/e3cfe2ce63a34382a2b23200cb9e0b1e" frameborder="0" width="1662" height="1246" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12461662Loomhttps://www.loom.com12461662https://cdn.loom.com/sessions/thumbnails/e3cfe2ce63a34382a2b23200cb9e0b1e-d4d4756543325425.gif539.095This Loom explains the shared relations import workflow and how imported node and relation schemas are matched or created. It describes importing shared relations in the DiscovererNode dialog by inverting external IDs, matching node schemas by existing imported identities or by schema ID and label case insensitively, and creating new discourse node schemas when no match is found. It then outlines relation schema matching, which is more complex: it reuses existing relation triples by source and destination when present, otherwise searches existing relation schemas that match the source and destination node types or creates a new relation schema. The speaker notes that enrichment of results is not performed during this step and mentions a planned change to only record imported source identity on creation.