Excessive "unknown" Nodes Created for Metadata Fields with Unknown Ontology IDs

When importing metadata nodes from cxg datasets, if the ontology ID for a metadata field is missing or marked as unknown/NA, the current implementation creates a new node for every occurrence. This behavior leads to the generation of numerous redundant nodes (each with a label/property value of "unknown") in the Knowledge Graph, which clutter the data and negatively impact query performance.