F21NL - Introduction to Natural Language Processing

Course leader(s):

Yannis Konstas (Edinburgh, September)

Aims

This course aims to:

lay the foundations of core concepts in Linguistics;
present the most common Natural Language Processing (NLP) problems together with appropriate Machine Learning solutions;
familiarise students with NLP applications in currently active research areas using Machine Learning techniques;
enable students to build simple NLP applications using commonly used libraries;
raise critical awareness of the appropriateness of NLP techniques and the relationships between them

Syllabus

1. Foundations in NLP (1.1 - Problems/Phenomena of interest in NLP, 1.2 - Machine Learning for NLP Primer, 1.3 - Distributional Semantics)

2. Machine Learning Architectures in NLP (2.1 1. Language Modelling with Neural Networks, 2.2 2. Sequence-to-Sequence Models, 2.3 3. Self-attention and Transformers, 2.4 4. Pre-trained Language Models)

3. NLP Applications (3.1 Popular NLP downstream tasks such as:, 3.2 1. Intent Prediction, Text Summarisation, Question Answering, 3.3 2. Beyond text-only: Caption Generation, Visual Question Answering)

Learning outcomes

By the end of the course, students should be able to do the following:

understand core concepts in Linguistics.
understand the linguistic motivation behind common Natural Language Processing (NLP) problems.
practice the mathematics underpinning methodologies of machine learning models for NLP.
differentiate between some popular machine learning modelling architectures to NLP.
assess the appropriateness and performance of different machine learning models to solve common NLP problems.
choose appropriate machine learning modelling libraries to conduct experiments for training and evaluating NLP systems.
conduct research independently to tackle practical problems.

Further details

Curriculum explorer: Click here

SCQF Level: 11

Credits: 15