Module 1: Introduction to Knowledge Graphs
Module 1: Introduction to Knowledge Graphs
1.1 What is a Knowledge Graph?
A Knowledge Graph (KG) is a graph-based data structure that represents knowledge as a set of entities
(nodes) and relationships (edges) between those entities.
Unlike traditional databases, a KG embeds semantics to provide richer context and meaning to the data.
Components of a Knowledge Graph:
- Entities (Nodes): Real-world objects or concepts (e.g., "Bill Gates," "Microsoft")
- Relationships (Edges): Connections that define how entities relate to each other (e.g., "Bill Gates is the
founder of Microsoft")
- Attributes (Properties): Descriptive information about entities and relationships (e.g., "Founded in 1975")
1.2 A Brief History of Knowledge Graphs
The concept of representing knowledge in graph form is not new. Its roots can be traced back to early
artificial intelligence research and semantic networks.
Timeline of Key Developments:
Module 1: Introduction to Knowledge Graphs
- 1972: The term "knowledge graph" was first used in research related to cognitive representation.
- 2012: Google introduces its Google Knowledge Graph, marking a major milestone in popularizing the
concept.
- 2013-Present: Major companies like Facebook, Microsoft, and Amazon adopt knowledge graphs for search,
recommendation systems, and personal assistants.
1.3 Why Are Knowledge Graphs Important?
Knowledge graphs provide:
1. Data Integration: Combine heterogeneous data from multiple sources.
2. Contextual Understanding: Relationships between entities add depth and meaning.
3. Reasoning and Inference: Use semantic rules to infer new facts.
4. Enhanced AI Systems: Improve the performance of AI applications in question answering,
recommendation systems, and natural language processing.
1.4 Structure of a Knowledge Graph
Module 1: Introduction to Knowledge Graphs
Knowledge graphs use graph-based models like:
1. RDF Graphs: Subject-predicate-object triples (e.g., "Bill Gates - founded - Microsoft").
2. Property Graphs: Support complex relationships with additional metadata on nodes and edges.
Schema and Ontologies:
A schema defines how data is structured in the graph, while an ontology defines the types of entities and
relationships.
Query Languages:
- SPARQL: For querying RDF-based knowledge graphs.
- Cypher: For querying property graphs (used in Neo4j).
1.5 Enhanced Comparison of KGs, Ontologies, KBs, and Semantic Networks
| Feature | Knowledge Graph (KG) | Ontology | Knowledge Base (KB) | Semantic Network
|-------------------------|-----------------------|------------------|---------------------|------------------------|
Module 1: Introduction to Knowledge Graphs
| Definition | Graph-based representation of entities and their relationships with formal semantics. |
Formal description of concepts and relationships. | Collection of structured facts and rules. | Network of
interconnected concepts.|
| Structure | Graph with nodes and edges enriched with semantics. | Hierarchical structure defining
classes and rules. | Set of facts (triples), rules, and mechanisms. | Directed graph with labeled edges.|
| Examples | Google Knowledge Graph, Wikidata | BioPortal Ontologies | MYCIN, Cyc | Conceptual
Dependency Theory|
1.6 Exercises
Exercise 1: Explore Existing Knowledge Graphs
1. Visit Wikidata and search for an entity of your choice.
2. Identify entities connected to the main entity, relationships, and properties.
Exercise 2: Build a Simple Knowledge Graph
1. Create a knowledge graph representing your academic background.
Module 1: Introduction to Knowledge Graphs
2. Use a tool like Graphviz or Neo4j to visualize your graph.
1.7 Recommended Reading
- Knowledge Graphs: A Comprehensive Introduction, Aidan Hogan et al.
- Your Guide to Knowledge Graphs by Diego Lopez Yse
- Building a Knowledge Graph to Enable Precision Medicine, PrimeKG.
