A Linked Data Model for Entity Aggregation

A fundamental requirement for the development of the Linked Data Knowledge Graph at the National Library Board of Singapore (NLB) was to able to ingest multiple entity descriptions, for the same real world object. Imagine all the individual Person entities for a prolific author that would be created if you import all the MARC records from a library management system.To satisfy this requirement an adaptive data model was developed utilising elements of the OAI-ORE Aggregation vocabulary to track the relationship between individual source derived entities (for the same object) and a primary entity which exhibits a consolidated view across all the sources.Richard will describe this simple model that is capable of dynamically tracking relationships even when source entities are deleted, modified, or added to the knowledge graph.