The University of Cyprus's course on Natural Language Processing is part of the Master programmes in Artificial Intelligence 4 Careers in Europe (MAI4CAREU). One of Master's programme's courses is split up into several lectures and is taught by Demetris Paschalides from the Department of Computer Science, whose research interest is focused on the use of Natural Language Processing (NLP) and Machine Learning (ML) to address social and ethical challenges.

Learning outcomes 

This lecture focuses on the application on NLP in political polarisation, providing case study examples from the news media to allow for a more comprehensive understanding of the topic. First, the lecture defines polarisation, and more specifically social polarisation. These are described as a major concern on a global scale with potential consequences in social cohesion and stability.

The lecture examines: 

• Polarisation Data Model 
• POLAR Framework (step-by-step process which includes:
  • Sentence segmentation
  • Named entity recognition and linking 
  • Generating sentiment attitude graph 
  • Identifying entity-to-entity relationships and calculating entity-to-entity attitude 
  • Extracting Entity Fellowships
  • Generating Fellowship Dipoles 
  • Dipole Topic Extraction 
  • Quantifying Topic Polarisation 
• POLAR Evaluation (comparing it with the PaCTE approach)

The lecture illustrates the theory with a lot of practical examples, illustrations, and interactive diagrams to offer students a comprehensive understanding of the course material. The instructor also provides a list of resources for further learning.

Course outline

The Natural Language Processing course is divided into 7 subsections: 

1. Text Pre-processing
2. Language Modeling
3. Text Classification
4. Word Vector Representation
5. Distributed Contextual Embeddings
6. Application of NLP in:
   1. Hate-speech Identification
   2. Fake News Detection
   3. Political Polarization
7. Introduction to Large Language Models