Knowledge Representation and Logic
Representing Knowledge
Why Representation Matters
An AI system cannot reason effectively unless it can encode facts about the world. Knowledge representation determines what the system can infer, explain, and update.
Logic-Based Reasoning
Propositional logic uses statements that are either true or false. First-order logic extends this with variables, predicates, and quantifiers, allowing more expressive statements about objects and relationships.
Representation Options
| Method | Strength | Weakness |
|---|---|---|
| Rules | Easy to understand | Difficult to scale |
| Graphs | Captures relationships | Can be ambiguous |
| Ontologies | Structured domain knowledge | Requires design effort |
| Probabilistic models | Handles uncertainty | Can be harder to interpret |
Closed World vs Open World
In some systems, what is not known is treated as false; in others, unknown information remains unknown. This assumption strongly affects inference.
What does first-order logic add beyond propositional logic?
First-order logic can express statements about objects and their properties more flexibly.
Correct answer: Variables, predicates, and quantifiers
Why are probabilities useful in AI reasoning?
Many real-world problems require reasoning with uncertain data rather than exact truth values.
Correct answer: They help model uncertainty when information is incomplete or noisy.
Symbolic vs Probabilistic Reasoning
Symbolic
- Uses explicit rules and logic
- Clear explanations
- Less tolerant of uncertainty
Probabilistic
- Represents uncertainty numerically
- Useful with incomplete data
- Often more computationally expensive
An ontology is best described as:
Ontologies organize domain concepts in a formal and structured way.
Correct answer: A structured description of concepts and relationships