392101 Knowledge Representation (S) (SoSe 2026)

The course covers foundational concepts and methods in Knowledge Representation and Reasoning (KR), addressing both theoretical foundations and practical applications. It focuses on symbolic and sub-symbolic approaches to representing, structuring, and reasoning over knowledge in intelligent systems.

Core topics include:

- First-Order Logic (FOL)

- Logic Programming and Automated Reasoning (Prolog programming, Rule-based reasoning)

- Structured Knowledge Representation (Frames, slots, and inheritance hierarchies)

- Start into Semantic networks

- Resource Description Framework (RDF)

- Knowledge graphs Reasoning

The second part of the course focuses on Knowledge Graph including:

- Knowledge acquisition from structured and unstructured data

- Distributional semantics and word embeddings

- Knowledge graph embeddings

- Graph Neural Networks (GNNs) for knowledge representation

Advanced application topics include:

- Knowledge tracing

- Parametric knowledge in large language models

- Knowledge-graph-enhanced recommender systems

Teilnahmevoraussetzungen, notwendige Vorkenntnisse

Useful prior knowledge: Discrete Mathematics, Formal Logic, Introduction to Machine Learning