• Assignment [30%]
• Formal examination [70%]

• Blackboard
• Lecturer's Module Page
• Unconfirmed practical marks when available
• Exam Papers, past 2 years (where applicable)

• to give students a well-rounded feel for the problems and approaches of Statistical Natural Language Processing (NLP)
• to give students an understanding of the potential areas of application of the techniques developed in Statistical NLP

• describe and discuss the subareas of NLP.
• implement NLP algorithms and techniques.
• describe and discuss the potential and limitations of NLP techniques for applications such as machine translation, question answering, and information extraction

Lectures will provide an overview of the field of NLP and its sub-areas, and will introduce and explain its key techniques, including their applicability and limitations. In lab classes, students will practice implementing the NLP techniques taught in class, testing their code in application to real language data. Topics covered will include:

• Introduction to NLP and its applications 
• Word embeddings 
• Language Modelling 
• Neural architectures for NLP: RNNs, Attention, Self-attention, and Transformers  
• Language Models: Tokenization, Pretraining, Encoding, and Decoding 
• Sequence Modelling: Sequence Tagging, Part-of-Speech Tagging, Hidden Markov Models (HMMs) and the Viterbi algorithm 
• Syntactic Parsing: Probabilistic Context-Free Grammars (PCFGs) and Dependency parsing 
• Recent Advances in NLP 
• Bias in NLP 

This module is only open to students who have taken Text Processing (COM3110/COM4115/COM6115) and Machine Learning and Adaptive Intelligence (COM4509/6509).

Optional modules within the school have limited capacity. We will always try to accommodate all students but cannot guarantee a place.