Inference & Knowledge Derivation Demo

Back to Lecture Overview

Understanding Inference

Key Distinction:

Entailment (Semantic)

KB ⊨ α

Meaning: α logically follows from KB in all possible worlds

Inference (Syntactic)

KB ├_i α

Meaning: α can be derived from KB using inference procedure i

Property	Soundness	Completeness
Definition	If KB ├_i α, then KB ⊨ α	If KB ⊨ α, then KB ├_i α
Meaning	Only derives valid conclusions	Derives all valid conclusions
Failure	Derives false conclusions	Misses valid conclusions

Interactive Inference Procedures

Explore different inference procedures and see how they derive new knowledge:

Modus Ponens

Sound Incomplete

Classic inference rule for implications

Resolution

Sound Complete

Complete inference method for propositional logic

Fallacious Rule

Unsound N/A

Example of unsound reasoning

Common Inference Rules

Sound Rules:

Modus Ponens (القياس المثبت): P→Q, P ├ Q

Definition: If P implies Q, and P is true, then Q must be true

Modus Tollens (القياس النافي): P→Q, ¬Q ├ ¬P

Definition: If P implies Q, and Q is false, then P must be false

And-Elimination: P∧Q ├ P

Or-Introduction: P ├ P∨Q

Unsound Rules (Fallacies):

Affirming Consequent: P→Q, Q ├ P ❌

Denying Antecedent: P→Q, ¬P ├ ¬Q ❌

Hasty Generalization: P(a) ├ ∀x P(x) ❌

Why Soundness & Completeness Matter

Soundness Protects Us

Prevents false conclusions from true premises
Essential for reliable automated reasoning
Maintains logical consistency
Builds trustworthy AI systems

Completeness Ensures Coverage

Finds all valid conclusions
Maximizes knowledge extraction
Critical for theorem proving
Enables comprehensive reasoning