Machine Intelligence with Knowledge Graphs.
A hybrid system generates diagnoses (Scenario) based on the patient’s symptoms and medical history stored in a Knowledge Graph. A human operator and an embeddings-based (ProcessingMethod) system (both Actors) with complementary scientific expertise (Capability) perform the prediction and justify their propositions to each other (InteractionTask) in order to come up with a final decision (EndGoal)
(dynamic) symptoms, embeddings, historical medical events, diagnoses, justifications, decisions. (static) human memory, medical KG
- (human) recollect similar events ;
- (human) induce the diagnosis;
- (machine) train and
- (machine) classify diagnoses;
- generate a justification for the diagnosis and
- idem
- present it to the other agent that can
- idem
- obtain the information;
- ultimately, generate a combined decision.
Prediction (1-4), Explaining (5-6), Team Role Allocation (7-10)