Data-driven semantic rule discovery for knowledge graphs using evolutionary computing

AI Knowledge Graphs (AI KGs) are realizations of Linked Enterprise Data, creating valuable business insights by bringing together semantic technologies and enterprise data infrastructures. These Knowledge Graphs go beyond data integration: the semantics underlying these next generation AI applications provide machines with the advantage of understanding the meaning of large data volumes. 
 
Knowledge Graphs are enriched by (semantic) axioms, rules and policies that allow for reasoning on data in the graph, in order to classify, find contradictions and create new insights on otherwise hidden relations. These rules are usually programmed and embedded in the semantic model (constraints) or added on top of the model (semantic rule logic). Knowledge systems in current production environments are typically pre-configured (model-driven) to apply rules and policies to data they contain. They must be (manually) enhanced or reconfigured to adapt to changes in their environment. 
 
We expect that more intelligent mechanisms for rule definition, with “cognitive” capabilities, will increase the adaptability and improve the applicability of these Knowledge Graphs. As new/unknown rules become apparent from data or rules change due to the dynamics of its environment, a Knowledge Graph can evolve and adapt its knowledge base and learn new rules and axioms as they emerge.