semantic networks编辑历史

# Semantic Networks

## 1 Definition and Core Concepts

Semantic networks are a form of knowledge representation that models relationships between concepts in a graph structure. In a semantic network, nodes represent concepts or entities, while edges (links) denote semantic relations such as "is-a," "part-of," or "causes." Originating in cognitive science and artificial intelligence, they enable machines to store, retrieve, and infer knowledge in a human-like associative manner. Semantic networks underpin many modern AI systems, including ontology-based reasoning, natural language processing, and knowledge graphs.

### 1.1 Basic Terminology

#### 1.1.1 Nodes and Edges
Nodes (also called vertices) represent discrete units of knowledge—concepts, objects, events, or attributes. Edges (links or arcs) connect pairs of nodes and represent directed or undirected binary relationships. The structure forms a directed graph when edges have a source and target, which is typical for most semantic networks.

#### 1.1.2 Labels and Properties
Both nodes and edges can carry labels that specify their meaning or type. Labels on edges indicate the relation (e.g., "is-a", "has-part"). Nodes may have associated properties (attributes) stored as key-value pairs, enabling richer representation beyond mere naming.

### 1.2 Fundamental Relations

#### 1.2.1 Taxonomic Relations (is-a, instance-of)
Taxonomic relations organize concepts into hierarchies. The "is-a" relation links a specific concept to a more general one (e.g., "Dog is-a Mammal"). The "instance-of" relation links an individual entity to its conceptual class (e.g., "Fido instance-of Dog"). These relations enable inheritance of properties.

#### 1.2.2 Partonomic Relations (part-of)
Partonomic relations describe composition or membership, e.g., "Wheel part-of Car" or "Finger part-of Hand." They support reasoning about wholes and parts, often with transitive properties.

#### 1.2.3 Causal and Temporal Relations
Causal relations (e