Natural Language Processing

Program Overview

This module offers an introduction to the current state-of-the-art techniques in the area of Natural Language Processing (NLP), namely Large Language Modules (LLMs). LLMs are among the defining technologies of our time. They form the core of modern Artificial Intelligence (AI) and are transforming how societies communicate and how knowledge is accessed.

Program Outcomes

This course introduces students to the foundations, architectures, and applications of modern language modeling. It covers both theoretical and practical aspects of LLMs, ranging from count-based and neural language models to Transformer architectures, alignment techniques, reasoning, retrieval-augmented generation (RAG), and multimodal extensions. A major emphasis is placed on developing ethical, transparent, and globally diverse models.

Learning Objectives

• Explain the fundamental principles of language modeling, including count-based, neural, and transformer-based approaches.
• Implement small-scale language models and conduct experiments using PyTorch and Hugging Face.
• Apply text generation techniques such as greedy search, temperature sampling, top-k, and nucleus sampling to control creativity and coherence in outputs.
• Integrate Retrieval-Augmented Generation (RAG) methods to enhance factual grounding and contextual relevance in model outputs.
• Develop and fine-tune specialized models, including reasoning models, multilingual models, and multimodal models that process text alongside images, audio, or video.
• Evaluate model performance using intrinsic and extrinsic metrics such as perplexity, BLEU, ROUGE, and human judgment.
• Understand and apply key techniques for LLM alignment, including instruction tuning, RLHF, and DPO.
• Design and prototype basic AI agents that use LLMs to perform tasks such as question answering, planning, and interaction with external tools.
• Identify, analyze, and mitigate ethical risks such as bias, misinformation, and under-representation in model behavior and datasets.
• Critically assess the capabilities, limitations, and societal impact of large language models in real-world applications.

Study Programs

This course is mandatory for the following study programs:

• MBA Digital Technologies
• MBA Digital Transformation

This course is offered as an elective for the following study programs:

• M.Sc. Applied AI
• M.Sc. Cybersecurity
• M.Sc. Advanced Digital Reality
• M.Sc. Digital Leadership

Micro Degree

• This course is offered as a micro degree.
• German UDS Micro Degrees are compatible with the European MOOC Consortiums Common Micro Credentials Framework.
• Micro Degrees will be rewarded with an equivalent of 5 ECTS.
• Micro Degrees are offered to non-regular students and require a fee of €900.

Requirements

None

General Information

• Teaching Format: Knowledge Essential
• Total Workload Master: 125h (40h/85h) / 5 ECTS
• Total Workload MBA: 100h (30h/70h) / 4 ECTS
• Total Workload Micro Degree: 125h (40h/85h) / Equivalent to 5 ECTS
• Module coordinator: Prof. Dr. Katya Artemova
• Examinations: Quizzes, presentation(s), essay(s)/paper(s), project report(s), written exam (tbd) - Details will be announced with course start.
• Offered: Odd quarters

Frequently Asked Questions

What web browser should I use?

Our German-UDS.academy platform works best with current versions of Chrome, Edge, Firefox, or Safari.

See our list of supported browsers for the most up-to-date information.