Backward Chaining

Backward Chaining is a goal-driven reasoning approach. Starting from a target hypothesis or goal, it identifies rules that can conclude this goal and recursively attempts to prove their preconditions as subgoals.

Category:Rule-Based Systems

•

Difficulty:Beginner

•

Time complexity:O(b^d)

•

Space complexity:O(d)

Interactive VisualizationBackward Chaining Visualizer

PHASE: INITIALIZEDGoal-driven reasoning (working backward from hypothesis)

Target Goal:

Knowledge Base (Rules)

Production rules structured as IF (antecedents) → THEN (consequent)

4 Rules

IFfeverANDcough

THENflu_suspected

IFflu_suspectedANDbody_ache

THENflu_likely

IFflu_likely

THENadvise_rest

IFrash

THENallergy

Working Memory

Known facts in the current state

0 Facts

Working memory is empty.

Goal Stack

Subgoals being verified

Active Stack (Top to Bottom)

Stack is empty.

0/0

Step Trace & Explanation-

Action:Click 'Run Inference' to start simulation.

Backward Chaining Pseudocode

1.prove(goal):

2. if goal in working memory: return true

3. for each rule that concludes goal:

4. if prove(every condition in rule.IF): add goal to memory, return true

5. return false (or ask the user)

Real-time Telemetry HUD

Live Stream

Nodes Explored

Frontier (Open Set)

Optimal Cost g(n)

—

Total Step Frames