Nexus Learn

Lecture 8: Inference and Resolution for Problem Solving

Lecture Slides CS-511 Artificial Intelligence 6 min read Updated 8 Nov 2026

↓ Download slides (PowerPoint)

Recap of Lecture 7 and Homework Check

LAST LECTURE · 5 MIN

Today: the six-step proof from L07 (S16) done by a machine - with ONE rule called resolution.

Today's Plan (50 minutes)

ROADMAP

0-5

Recap of Lecture 7 and homework check

S02

5-8

Plan and learning outcomes

S03-S04

8-16

Module 1: Inference rules and proofs

S05-S08

16-27

Module 2: The resolution rule and CNF

S09-S11

27-35

Module 3: Proof by resolution; unification

S12-S15

35-43

Class activity: prove it

S16

43-48

Common mistakes and recap quiz

S17-S18

48-50

Homework and next lecture

S19

Learning Outcomes (SLOs)

BY THE END OF THIS LECTURE YOU WILL BE ABLE TO

SLO 8.1

Apply inference rules (modus ponens, modus tollens and others) to derive a conclusion

→ CLO2 · C3 Apply

SLO 8.2

Convert a propositional sentence into conjunctive normal form (CNF)

→ CLO2 · C3 Apply

SLO 8.3

Prove a query by resolution refutation and unify simple first-order atoms

→ CLO2 · C3 Apply

  • Course Learning Outcome(s) served
  • CLO2: Demonstrate first-order predicate calculus, logical reasoning and problem-solving using Prolog language.

Slide 5

MODULE 1

Inference Rules

Proving new sentences from old ones - without a truth table

What Is Inference?

MODULE 1 · INFERENCE

Truth tables

Inference rules

Sure, but 2ⁿ rows: 50 atoms give about 10¹⁵ rows.

Work on the sentences: a proof of 10-20 steps, whatever the number of atoms.

The Main Rules of Inference

MODULE 1 · RULES

Why sound: each rule is a tautology (L06) - it can never lead from true premises to a false conclusion.

A Proof with Inference Rules

MODULE 1 · PROOF

The KB

  • R1: F ∧ B → V
  • R2: V ∧ S → T
  • R3: T ∧ L → D
  • Facts: F, B, S, L
  • F fever, B body ache, V viral, S dengue season, T test, L low platelets, D dengue suspected

Problem: which rule to try next? A computer needs ONE simple rule it can always apply. That rule is resolution.

Slide 9

MODULE 2

Resolution and CNF

One rule that does the work of all the others

The Resolution Rule

MODULE 2 · RESOLUTION

From A ∨ P and ¬P ∨ C conclude A ∨ C

Words: literal = P or ¬P; clause = literals joined by ∨; resolvent = the new clause; □ = empty clause (contradiction).

Converting to Conjunctive Normal Form (CNF)

MODULE 2 · CNF

REAL OUTPUT: L08/CNF.PY (CHECKED IN ALL 8 ROWS)

W ↔ (C ∧ ¬R) vs (¬W ∨ C) ∧ (¬W ∨ ¬R) ∧ (¬C ∨ R ∨ W) EQUIVALENT

Result: 3 clauses: ¬W ∨ C, ¬W ∨ ¬R, ¬C ∨ R ∨ W. An implication A ∧ B → C becomes one clause ¬A ∨ ¬B ∨ C.

Slide 12

MODULE 3

Proof by Resolution

To prove α, assume NOT α and reach a contradiction

Resolution Refutation: The Method

MODULE 3 · REFUTATION

  • 1. Write the KB in CNF (a list of clauses).
  • 2. Add the NEGATION of the query: to prove α, add ¬α.
  • 3. Resolve pairs of clauses and add each new resolvent.
  • 4. Empty clause □ found → contradiction → the query is PROVED.
  • No new clauses possible → the query does NOT follow.

REAL OUTPUT: L08/RESOLUTION.PY

  • Example 1: KB = {A → E, ¬E}. Query: ¬A
  • 1. A negated query
  • 2. ¬A ∨ E KB
  • 3. E resolve 1 + 2
  • 4. ¬E KB
  • 5. □ (empty clause) resolve 3 + 4
  • The empty clause was reached, so the query is PROVED.

Example: L06 homework Task 3 (attendance) - modus tollens, now done by resolution.

Demo: A Resolution Prover in Python

MODULE 3 · REFUTATION

REAL OUTPUT: EXAMPLE 2 - DENGUE KB

REAL OUTPUT: EXAMPLE 3

  • Example 2: dengue KB (3 rules, 4 facts). Query: D (dengue suspected)
  • 1. ¬D negated query
  • 2. D ∨ ¬L ∨ ¬T KB
  • 3. ¬L ∨ ¬T resolve 1 + 2
  • 4. ¬S ∨ T ∨ ¬V KB
  • 5. ¬L ∨ ¬S ∨ ¬V resolve 3 + 4
  • 6. ¬B ∨ ¬F ∨ V KB
  • 7. ¬B ∨ ¬F ∨ ¬L ∨ ¬S resolve 5 + 6
  • 8. F KB
  • 9. ¬B ∨ ¬L ∨ ¬S resolve 7 + 8
  • 10. B KB
  • 11. ¬L ∨ ¬S resolve 9 + 10
  • 12. S KB
  • 13. ¬L resolve 11 + 12
  • 14. L KB
  • 15. □ (empty clause) resolve 13 + 14
  • The empty clause was reached, so the query is PROVED.
  • Example 3: KB = {R → W, W}. Query: R (converse trap)
  • no proof: the query does not follow

Clauses in

  • ¬F ∨ ¬B ∨ V, ¬V ∨ ¬S ∨ T, ¬T ∨ ¬L ∨ D, F, B, S, L and the negated query ¬D.
  • Strategy: set of support - every step uses the negated query or a clause made from it.

Resolution in First-Order Logic: Unification

MODULE 3 · FIRST-ORDER LOGIC

REAL OUTPUT: L08/UNIFY.PY

  • UNIFY(Eligible(x), Eligible(Hira)) = {x/Hira}
  • UNIFY(Knows(Ali, y), Knows(x, Sana)) = {x/Ali, y/Sana}
  • UNIFY(Advises(Waqas, x), Advises(y, fatherOf(Ali))) = {y/Waqas, x/fatherOf(Ali)}
  • UNIFY(Knows(Ali, y), Knows(y, Sana)) = FAIL (no substitution)
  • UNIFY(Student(Ali), Teacher(Ali)) = FAIL (no substitution)

Prolog (L09): eligible(X) :- student(X), att75(X). ?- eligible(ali). Prolog unifies X = ali and resolves.

Activity: Prove It

CLASS ACTIVITY · PAIRS · 8 MIN

Pairs, 6 minutes + 2 minutes check. Write the answer and the rule you used.

Name the rule: 'If it rains the match is cancelled. It rains. So the match is cancelled.'

Convert ¬(A ∧ B) → C to CNF.

Name the rule: from R → C and ¬C conclude ¬R.

Resolve the clauses A ∨ B and ¬B ∨ C.

Convert P → (Q ∨ R) to CNF.

Prove R by resolution from KB = {P → Q, Q → R, P}.

Common Mistakes

WATCH OUT

WRONG IDEA

CORRECT IDEA

✗ To prove α, add α to the KB.

✓ Add ¬α (the NEGATION) and look for the empty clause.

✗ From P → Q and Q, conclude P.

✓ Invalid (converse trap). Only modus ponens (P) and modus tollens (¬Q) are valid.

✗ Resolve two pairs at once: (P ∨ Q), (¬P ∨ ¬Q) gives □.

✓ Resolve ONE pair: the result is Q ∨ ¬Q (a tautology), not □.

✗ A → B becomes A ∨ ¬B in CNF.

✓ A → B ≡ ¬A ∨ B (negate the left side).

✗ Knows(Ali, y) unifies with Knows(y, Sana).

✓ It fails: y cannot be Ali and Sana. Rename variables apart first.

Recap Quiz

CHECK YOUR UNDERSTANDING · 3 MIN

  • Q1. From P → Q and ¬Q we conclude ¬P. This is:
  • (a) Modus ponens (b) Modus tollens (c) AND-elimination (d) Resolution only
  • Q2. Resolving (A ∨ B) with (¬B ∨ C) gives:
  • (a) A ∨ C (b) A ∧ C (c) B ∨ C (d) □
  • Q3. In CNF, A → B becomes:
  • (a) A ∨ ¬B (b) ¬A ∧ B (c) ¬A ∨ B (d) A ∧ ¬B
  • Q4. To prove α by refutation we add:
  • (a) α (b) ¬α (c) α ∨ ¬α (d) Nothing

Homework and Next Lecture

BEFORE THE NEXT CLASS

  • Task 1: Convert to CNF: (a) (L ∧ ¬U) → ¬F (b) S ↔ E.
  • Task 2: D = soil dry, R = rain forecast, I = irrigation needed, L = load-shedding, G = start diesel engine. KB = {(D ∧ ¬R) → I, (I ∧ L) → G, D, ¬R, L}. Prove G by resolution (number every step).
  • Task 3: Find the unifier or say FAIL: (a) Teaches(Waqas, x) and Teaches(y, CS511) (b) Older(x, x) and Older(Ali, Sana).
  • Read: Russell & Norvig 7.5 and 9.2; Luger 14.1. Install SWI-Prolog (swi-prolog.org) for Lecture 9.

NEXT LECTURE

  • Lecture 9:
  • Logic Programming with Prolog

← All CS-511 material