EN 
Українська
PolskiExpert System Inference Simulator (2D)
2D rule network for a classic expert system: watch forward chaining fire facts into rules toward a conclusion, or trace backward chaining from a goal to the facts that support it, with rule confidence and chain speed tunable in real time.
This 2D companion drives the same forward/backward chaining inference engine as the 3D version through a flat canvas view: base facts sit in one column, the rules that consume them sit in the next, and derived facts feed the following column up to a final diagnosis node. Run forward chaining to watch evidence ignite rules and propagate toward a conclusion, or switch to backward chaining to see the engine work backward from the goal, verifying each subgoal against the facts and rules that support it — with a tunable rule-confidence slider that can stall the chain even when the underlying facts are true, modelling the uncertainty real diagnostic expert systems must tolerate.
2D rule network for an expert system: forward chaining fires facts into rules toward a conclusion, backward chaining traces a goal back to its supporting facts, with rule confidence and chain speed tunable in real time.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install
Forward chaining starts from known facts and fires every rule whose antecedents are satisfied, propagating new facts forward until a conclusion is reached. Backward chaining starts from a goal and works backward, checking whether the rules and facts needed to prove it are available.
Rule confidence is the probability that a satisfied rule actually fires, modelling the uncertainty real expert systems must tolerate. At low confidence, a rule whose antecedents are true can still fail to fire, which can stall the whole inference chain.
The chain stalls when no rule can fire toward the goal — either because the symptom scenario leaves a required antecedent fact false, or because a satisfied rule was blocked by the random confidence check.