Patterns in large real-world datasets

Discover meaningful patterns in data to tell us how the world is (or was in the past), so as to understand the past, present and future better, as a foundation for a critical response, and to argue for change. Data may originate from many sources and in various formats, including texts, images, videos, audios. Databases, AI, and coding are key capabilities for this.

Examples

Jinhua Yang (WMG) and Geoffrey Allan Rhodes (Shanghai Jiao Tong University) collect and analyse (using AI) audience responses to immersive arts performances, using large datasets harvested from social media. They look for factors that lead to significant differences in people’s responses. This could inform changes in practice, for example by identifying how artists can engage better with under-represented audiences.

Technologies and techniques

  • Integrated development environments (Visual Studio Code, R Studio).
  • Relational databases and Structured Query Language (SQL).
  • Procedural programming/scripting (Python, R).
  • Object databases and query tools (JavaScript, node.js).
  • Text-based collections, search and analysis tools (corpus linguistics, sentiment analysis, AI, Gale Digital Scholar).
  • Geo data and geographical information systems (GIS).
  • Social media data mining and analysis.

Next

Patterns in large real-world datasets | Human stories in close-up | Creative practice and experimentation

We also run (and record) sessions that cover transferable capabilities.