Research Group: Machine intelligence

About our research

Our group includes:

• the computational intelligence laboratory
• a special interest group on artificial intelligence (AI) and data ethics
• a special interest group on computer science and AI for education

Computational intelligence laboratory

Our research covers neural networks, fuzzy systems, clustering and evolutionary computation.

We use biology-inspired methods to develop new theories and algorithms for machine learning and data mining.

Our focus is to:

• perform internationally recognised and world-class research in the analysis, design and application of computational intelligence methodologies and algorithms with positive societal impact
• transfer knowledge and expertise to industry across diverse sectors
• conduct research following ethical artificial intelligence guidelines and practices and consider the moral, social, environmental and legal implications of emerging technologies
• engage the public in our research activities through co-creation and co-production activities
• upskill our students through research-informed teaching and project work

Our research areas include:

• adaptive psychological profiling ⁠— using machine learning to detect human mental states such as deception, comprehension and guilt from non-verbal behaviour (examples include Silent Talker, Fathom and COMPAS)
• chatbots ⁠— applications include student debt advice, museum guides, special needs teaching, and conversational intelligent tutoring which is adaptive to comprehension levels
• semantic similarity of natural language ⁠— focusing on English and low resource languages (Arabic, Urdu and Thai) to develop word and short text similarity measures and fuzzy natural language processing to improve communication with machines (our novel measures have also been used to develop new semantic clustering algorithms and have been successfully applied to Twitter data)
• theoretical and technical algorithm development and applications ⁠— including artificial neural networks, trees, computing with words, fuzzy trees and ensembles
• nature-inspired algorithms and games ⁠— including ant colony optimisation
• fuzzy systems ⁠— such as road congestion management and fuzzy decision trees for decision making
• smart data ecosystems ⁠— for the detection of irregular behaviour in financial systems

AI and data ethics

Our multidisciplinary team works with businesses on knowledge exchange and policy design, exploring responsibility, accountability, bias, transparency, fairness, trust, explainability and AI governance.

We examine the legal, social, moral, environmental and ethical concerns of AI services and systems in education, healthcare and business. Our team recently contributed to the Public Intelligence whitepaper: Empowering Accountable and Ethical Innovations.

Using co-design and co-production strategies, we work with businesses to provide education and skills to bridge the knowledge gap and empower the public in AI and data ethics.

Computer science and AI for education

We explore the application of computer science ⁠— and more specifically AI ⁠— in education.

Our team conducts research into conversational intelligent tutoring systems and new systems for assessing the programming language SQL at undergraduate level.

We place a strong emphasis on the ethical implications of the use of AI in the general field of education.

Selected knowledge transfer projects

We work with a range of businesses on knowledge transfer projects, including:

• The Insights People (2020 to 2023) ⁠— developing the use of semantic-based clustering algorithms for consumer profile surveys
• Access Pay (2018 to 2020) ⁠— embedding a machine learning capability to enhance financial forecasting solutions and support growth,  awarded outstanding by Innovate UK
• Service Power Ltd (2015 to 2019) ⁠— developing a cloud-based staff rostering product that integrates with field service applications, awarded outstanding by Innovate UK
• Service Power Ltd (2014 to 2016) ⁠— advancement of their technological base by applying new optimisation algorithms to service offerings