Natural Language Processing and Large Language Models module (CS52003)

Language is central to how we communicate, learn, and share knowledge, and teaching machines to understand it has become one of the most exciting challenges in artificial intelligence. This module provides a practical and theoretical foundation in Natural Language Processing (NLP), introducing the methods that allow computers to interpret, analyse, and generate human language.

You’ll learn how the field has evolved from early probabilistic language models to today’s transformer-based systems that power large language models (LLMs) such as GPT. Through hands-on work, you’ll experiment with pre-trained models, fine-tuning them for real-world tasks like classification, summarisation, and question answering. You’ll also examine ethical issues including bias, fairness, and the responsible use of large-scale AI systems.

What you will learn

In this module, you will:

• explore foundational concepts and techniques in Natural Language Processing
• study how models learn to represent and generate text through tokenisation, embeddings, and attention mechanisms
• learn how to pre-train, fine-tune, and evaluate large language models for specific tasks
• apply NLP techniques to real-world challenges such as text classification, summarisation, and question answering
• examine key ethical, fairness, and computational considerations in building and deploying LLMs

By the end of this module, you will be able to:

• explain and apply the core concepts of modern NLP and deep learning architectures
• develop and adapt NLP systems using existing frameworks and libraries
• evaluate the performance and limitations of large language models in applied contexts
• critically assess ethical and societal challenges in the use of language-based AI systems

Assignments / assessment

• research project (50%)
• practical project (50%)

Both components focus on the development and evaluation of NLP systems, combining research insight with hands-on experimentation.

This module does not have a final exam.

Teaching methods / timetable

You will learn through a mix of lectures, workshops, and practical lab sessions.
Workshops and labs will involve guided experiments with real datasets and models, encouraging independent exploration and critical reflection on the capabilities and limitations of LLMs.