Cookies

We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Faculty Handbook Archive

Archive Module Description

This page is for the academic year 2021-22. The current handbook year is 2022-23

Department: Computer Science

COMP4167: NATURAL LANGUAGE PROCESSING

Type Open Level 4 Credits 10 Availability Available in 2021/22 Module Cap None. Location Durham

Prerequisites

  • COMP3547 Deep Learning and Reinforcement Learning OR COMP2231 Software Methodologies

Corequisites

  • None

Excluded Combination of Modules

  • None

Aims

  • Introduce the students to computational linguistics
  • Introduce the students to Language models
  • Gain experience in working with various textual data
  • Gain experience in using advanced techniques to solve natural language tasks such as text parsing, understanding, classification, translation, and generation

Content

  • Text Pre-processing
  • Features Extraction
  • Language Modelling and Neural Language Modelling
  • Neural Word Embedding and their interpretation
  • Recurrent Neural Networks (RNN) for Language Models
  • Advanced variations of RNNs
  • Sequence to Sequence Architectures
  • Convolutional Neural Networks for Text Classification
  • Transformers and Attention based Models
  • Multitask Learning
  • Natural Language Generation
  • NLP Ethics and Fairness

Learning Outcomes

Subject-specific Knowledge:
  • On completion of the module, students will be able to demonstrate:
  • Understanding of the fundamental concepts of Language Models
  • Understanding of the mathematical basis of various deep-learning based language models
  • Understanding of the learning algorithms behind major NLP use cases e.g. Machine Translation, Multi-task Learning, Language Generation, ...
  • Understanding of the embedded bias in popular language models
Subject-specific Skills:
  • On completion of the module, students will be able to demonstrate:
  • The ability to conduct independent research in the NLP field
  • The ability to handle textual data and extract representative features
  • The ability to use state-of-the-art NLP techniques and models to solve real-world applications
Key Skills:
  • On completion of the module, students will be able to demonstrate:
  • The ability to design end-to-end solutions for real-world problems with textual input using state-of-the-art NLP techniques
  • The ability to make informative decisions regarding the Deep Learning choices and the word embeddings
  • Awareness of the Language Models and their biases

Modes of Teaching, Learning and Assessment and how these contribute to the learning outcomes of the module

  • Lectures enable the students to learn new material relevant to NLP concepts, word embeddings, language models, as well as their applications.
  • Practical classes enable the students to put into practice learning from lectures and strengthen their understanding through application.
  • Summative assessments assess the application of methods and techniques, and assess the understanding of core concepts.

Teaching Methods and Learning Hours

Activity Number Frequency Duration Total/Hours
lectures 22 2 per week, unless there is a practical class that week 1 hour 22
practical classes 2 2 set within the teaching period of the module 1 hour 2
preparation and reading 76
total 100

Summative Assessment

Component: Coursework Component Weighting: 100%
Element Length / duration Element Weighting Resit Opportunity
Summative Assignment 100% NO

Formative Assessment:

Example formative exercises are given during the course.


Attendance at all activities marked with this symbol will be monitored. Students who fail to attend these activities, or to complete the summative or formative assessment specified above, will be subject to the procedures defined in the University's General Regulation V, and may be required to leave the University