Date: 08/10/2026
Masterclass: Analysing large volumes of data with AI and computational methods
At a glance
When: Friday 30 April 2027, 9.30 am to 4.00 pm
Where: University of Southampton, room to be confirmed
Who it is for: Doctoral students from qualitative, quantitative or mixed-methods backgrounds who want to analyse large collections of data using accessible AI tools.
Focus: Specialist; Methods
Overview
Social scientists increasingly find themselves holding more data than they can read: thousands of documents, transcripts, posts or records. This masterclass looks at what AI and computational methods can realistically do with material at that scale, and – just as importantly – what they cannot.
The day comprises two practical workshops, each preceded by a short theory lecture. Lunch is provided.
Training focus
Session 1: Working with Large Numerical Datasets
From exploration to prediction, this session focuses on structured data (tables, spreadsheets, survey responses, administrative records and similar). It covers practical, scalable workflows for managing and cleaning larger datasets without software crashes, exploring key trends, feature engineering, and applying supervised machine learning models.
Skills you will gain:
Hands-on task: work with a real-world tabular dataset in Python to build, tune and evaluate a predictive classification model on structured data.
Session 2: Analysing Large Volumes of Text and Images with AI
This session focuses on unstructured data (policy documents, reports, news archives, open-ended surveys and image datasets), covering modern computational tools for organising and summarising text and visual materials at scale. You will learn how to automatically discover topics across thousands of documents, search text and image libraries using AI embeddings, summarise open-ended survey data, and use computer vision to categorise image sets.
Skills you will gain:
Hands-on task: analyse a large document corpus or image set to generate automated themes, run semantic searches, and compare AI-generated categories against human qualitative judgment.
What you’ll need
Session 1: Python via Google Colab, using standard libraries such as pandas, scikit-learn and XGBoost.
Session 2: Python tools in Google Colab, such as BERTopic, sentence-transformers and pre-trained Hugging Face models.
Facilitator
Dr Asra Aslam
Lecturer in Data Science, University of Sheffield
Dr Aslam specialises in AI, machine learning, computer vision, and their applications in health sciences. She has led projects on AI-driven healthcare, collaborated with industry and public bodies such as Transport for London, and received multiple national and international awards recognising her contributions to AI, health, and equity, diversity and inclusion.
01 / SCDTP programme
SCDTP Training
Training organised and funded by the SCDTP.
