Knowledge Representation and Reasoning

Knowledge Representation and Reasoning

Knowledge representation (KR) encodes information about the world in a form that AI systems can use for reasoning. Good KR enables inference, explanation, and intelligent decision-making.

Propositional Logic

Propositional logic uses boolean variables and connectives (AND, OR, NOT, IMPLIES). It supports truth tables, logical equivalence, and inference rules (modus ponens, resolution). Limited because it cannot express relationships between objects or quantify over sets.

First-Order Logic

First-order logic (FOL) adds objects, predicates, functions, and quantifiers (for-all, there-exists). Example: ForAll x: Student(x) IMPLIES Person(x). FOL is more expressive than propositional logic. Inference methods: forward chaining, backward chaining, and resolution.

Semantic Networks

Semantic networks represent knowledge as graphs with nodes (concepts) and edges (relationships). IS-A links represent inheritance. Advantages: intuitive visualisation. Limitations: no standard semantics, difficulty with exceptions. Knowledge graphs (Google, Wikidata) are modern semantic networks.

Frames

Frames (Minsky) represent stereotypical situations as structured objects with slots and fillers. Similar to object-oriented classes with default values and inheritance. Frames capture expectations and enable reasoning about typical situations and exceptions.

Ontologies

An ontology formally defines concepts, relationships, and rules for a domain. OWL (Web Ontology Language) is the standard. Ontologies enable knowledge sharing, interoperability, and automated reasoning. The Semantic Web uses ontologies for machine-readable web content.

Expert Systems

Expert systems apply domain knowledge to solve specific problems. Architecture: knowledge base (rules), inference engine (reasoning), working memory (current facts), explanation facility, and user interface. Examples: MYCIN (medical diagnosis), DENDRAL (chemistry). Rule-based reasoning uses IF-THEN rules.

Summary

Knowledge representation — logic, semantic networks, frames, ontologies, and expert systems — enables AI to store, reason about, and apply knowledge. KR is fundamental to building intelligent systems that explain their reasoning.