2511606 – Knowledge-driven Artificial Intelligence

This master-level course takes you beyond the hype of modern AI and into the deeper question that defines the next frontier: not just how machines learn, but how they know. Bridging the gap between raw data and genuine intelligence, Knowledge Driven AI equips you with a rare double fluency — in the statistical power of machine learning and the structured precision of symbolic knowledge representation. You will trace the full journey from numbers to meaning: starting with the foundations of information theory and human communication, moving through the mechanics of machine learning and neural networks, and diving deep into the rich world of Natural Language Processing. Along the way, you'll discover how language models work, how words carry meaning, and how machines can be taught to read. But the real adventure begins with Knowledge Graphs — the backbone of the modern semantic web. You'll learn to model the world using ontologies, query it with SPARQL, and validate it with SHACL, working hands-on with real-world graphs like Wikidata and DBpedia. The course culminates in Neurosymbolic AI — where logic meets learning — giving you a front-row seat to one of the most exciting and unresolved debates in the field: Can neural networks ever truly reason?
Whether your passion lies in building smarter NLP pipelines, designing knowledge-rich applications, or pushing the boundaries of what AI can do, this course gives you the conceptual depth and practical toolkit to do it thoughtfully — and brilliantly.

1) The Art of Understanding
- From Numbers to Insights
- Data, Information, and Knowledge
- Natural Language and Successful Communication
- The Art of Understanding
- The Principles of Learning

2) Basic Machine Learning
- Machine Learning Fundamentals
- How to evaluate a Machine Learning Experiment?
- Evaluation and Generalisation Problems
- Supervised and Unsupervised ML
- Basic ML Algorithms
- Linear Regression
- Decision Trees
- k-means Clustering
- Neural Networks and Deep Learning

3) Natural Language Processing
- NLP and Basic Linguistic Knowledge
- NLP Applications, Techniques and Challenges
- Regular Expressions, Tokenisation and Word Normalisation
- Statistical Language Models (N-Gram Model)
- Naive Bayes Text Classification
- Distributional Semantics and Neural Language Models
- Word Embeddings

4) Knowledge Graphs
- Knowledge Representations and Ontologies
- Resource Description Framework (RDF)
- Modeling with RDFS
- Querying RDF(S) with SPARQL
- Popular Knowledge Graphs - Wikidata and DBpedia
- Ontologies with the Web Ontology Language (OWL)
- Linked Data Quality Assurance with SHACL
- The Graph in Knowledge Graphs

5) Neurosymbolic AI
- Symbolic and Subsymbolic AI
- Knowledge Graph Embeddings and KG Completion
- The Limits of AI
- KDAI Master Thesis Topics

Learning objectives:
- The students know the fundamentals and measures of information theory and are able to apply those in the context of Knowledge-driven AI.
- The students have basic skills of natural language processing and are enabled to apply natural language processing technology to solve and evaluate simple text analysis tasks.
- The students have fundamental skills of knowledge representation with ontologies as well as basic knowledge of Semantic Web and Linked Data technologies. The students are able to apply these skills for simple representation and analysis tasks.
- The students know the fundamentals of basic machine learning technologies. The students are able to apply these skills for simple prediction and classification tasks.
- The students know the basic principles of neurosymbolic AI.

Englisch

Artificial Intelligence, Machine Learning, NLP, Knowledge Graphs, Neurosymbolic AI

All rights reserved

Unbegrenzt – wenn online geschaltet

Sie können diesem Kurs direkt beitreten.

Bis: 30. Apr 2026, 17:20

Anmeldename
Vorname
Nachname
E-Mail
Matrikelnummer

3819530