Akitaka Matsuo is a Lecturer in the Department of Government at the University of Essex. Prior to his current role, he was a Postdoctoral Fellow at the Institute for Analytics and Data Science (IADS). Before joining IADS, he was a Research Fellow in Data Science in the Department of Methodology at the London School of Economics and Political Science. He received his PhD in Political Science from Rice University in 2012. For more information about his research and publications, please visit his website: https://amatsuo.net

Course Description
Gain a practical understanding of artificial intelligence (AI) and large language models (LLMs), and learn how to apply them for text analysis, data extraction, and responsible decision-making. This course is designed for policymakers and analysts looking to integrate AI tools into data-driven work. There are no prerequisites, but basic familiarity with R (as covered in Module 1) is required, and the more advanced skills introduced in Module B are beneficial. No prior experience with machine learning or natural language processing is necessary.

Module 1: Foundations of AI and LLMs in Data Science
Gain a conceptual and practical understanding of artificial intelligence and large language models, exploring how they learn, generate text, and integrate into modern data-science workflows.

Module 2: Text as Data with LLMs: Classification, Extraction & Summarisation
Learn how to use LLMs to transform unstructured text into structured data for analysis, applying techniques such as classification, sentiment detection, and summarisation to policy-relevant texts.

Module 3: AI as a Partner in Data-Science Work
Explore how AI can act as a collaborator in data-analysis workflows from generating and validating code to building secure, reproducible, and responsible analytical pipelines.

Applications open soon!

Course Outline

Module 1: Foundations of AI and LLMs in Data Science
Gain a conceptual and practical understanding of artificial intelligence and large language models, exploring how they learn, generate text, and integrate into modern data-science workflows.

  • Non-technical overview of large language models (LLMs)
  • Introduction to core machine-learning concepts and workflows
  • Handling text in R
  • Using LLMs locally and through API from R

 

Module 2: Text as Data with LLMs: Classification, Extraction & Summarisation
Learn how to use LLMs to transform unstructured text into structured data for analysis, applying techniques such as classification, sentiment detection, and summarisation to policy-relevant texts.

  • Working on various text-analytics tasks with LLMs:
    • Sentiment analysis
    • Stance detection
    • Scaling
    • Topic classification
  • Information extraction (e.g., pulling names, organisations, topics from reports)
  • Summarisation of documents with LLMs

 

Module 3: AI as a Partner in Data-Science Work
Explore how AI can act as a collaborator in data-analysis workflows from generating and validating code to building secure, reproducible, and responsible analytical pipelines.

  • AI-assisted coding and analytics: generating scripts, cleaning data, exploratory analysis, interpretation
  • Learn to design, critique, and validate AI-assisted data projects
  • Setting up customised AI-assisted data-analysis workflows
  • Discussing security, reproducibility, and responsible governance of AI-assisted work