Applied Data Analytics Group

Vast amounts of data are generated every day: from the text and visual content placed online by users and providers, to the click trails and other collateral data emanating from users’ interaction with the online world, we need better methods to make sense of, manage, organise, visualise, utilise and monetise data.

The common aim in all data analytics is to turn data into information, and ultimately knowledge, that can be used to add value to businesses, enrich user experiences, help people lead healthier lives at all ages, and fast-track design processes.

The two sides in a data conversation are (1) owners/producers of data, i.e. people with data and questions about it; and (2) data analysts, i.e. people with the expertise to answer such questions, and, often more importantly, to identify the best way to ask them.

The Brighton Data Analytics Group at the University of Brighton aims to make it easier for data owners/producers and data analysts to start a conversation, and to make it more likely that such conversations lead to tangible results and demonstrable impact. In response to a growing interest in, and demand for, big data and data analytics within the university and across virtually all economic sectors, we have formed a group of computer scientists, statisticians and mathematicians with complementary expertise in data analytics, in particular in statistical analysis and machine learning, and with a track record of applying specific data analysis techniques to real-world numerical, language and image data.


Group members' expertise is grounded in new and established methods and tools from computer science, statistics and mathematics including a wide range of different machine learning and statistical modelling techniques. Building on these are more specialised approaches for the analysis of textual, image and structured numerical data. We have worked with a wide range of different specific data types, including:

  • Social media posts
  • Free-text comments
  • Medical images
  • User-generated photographs
  • Geolocation data
  • Corporate and financial data
  • Medical data

Our published data analytics research involving the above data types includes:

  • Scanning of Twitter posts to detect mentions of given drugs and their effects in the context of post-marketing surveillance for pharmacovigilance
  • Automatic analysis and description of entities in images and relations between them
  • Automatic understanding of 3D data in accordance with human visual perception
  • 2D and 3D pattern analysis with Markov/Conditional Random Fields
  • Reconstruction of topological structures (such as the structure of a planar graph) from point cloud data
  • Forecasting the number of answers to questions posted on websites
  • Automatic selection of side information (context) in forecasting implied volatility and student test performance
  • Analysis of user survey studies on diagrammatic reasoning
  • Analysis of gene-expression to model intra-tumour variability in cancer studies
  • Modelling association between functional, epigenetic, and expressional profile in limb ischemia
  • Analysis of longitudinal studies


Depending on the complexity and confidentiality of the task at hand, we engage with external and internal partners in one of several collaboration modes:

  1. Smaller-scale projects, funding not always required:
    • 3rd year UG and MSc projects in maths and computing: selected individual UG and MSc students supervised by ADA group members
    • 3-month summer internship projects: carried out by a team of summer interns supervised by ADA group members
  2. Research collaborations, third-party funded:
    • PhD projects
    • research projects
  3. Industry-oriented projects:
    • student placements
    • KTPs
    • consultancy work