LIASD brings together computer science lecturers and doctoral students, and spans three research themes studied by EID, IUSD and PASTIS research groups.

EID or the Smart Data Spaces group specialises in Big Data engineering. Its work is underpinned by the Unit’s skills and experience in the field of data and knowledge. In this context, the research work is now split into two fields of expertise: deductive systems that produce knowledge through reasoning using symbolic models, and inductive systems that produce knowledge from Big Data through statistical and neural learning. Based on our studies, recent advances in these areas of research, new challenges, recent calls for projects and our contacts with industry and academia, we believe that developing hybrid systems is the right path. As a result, we propose avenues of research that allow for different types of hybrid models and systems to be created, integrating inductive learning into deductive systems and vice versa to varying degrees. In this theme, we also look at personal data protection, which is now a central issue in the field. In order to further characterise the research direction we propose to follow in the coming years, below we describe our main research orientations. These projects are organised into five areas:

1. Deductive reasoning on knowledge and data
2. Inductive learning using Big Data
3. Dynamic multi-modal models
4. Multilingual semantic resources
5. Formal approach to personal data

The IUSD research group is an abbreviation of Ubiquitous Computing and Data Science. The work of the IUSD group is dedicated to the development of algorithms and methodologies for human-centred or human-in-the-loop applications. It explores the themes of biometrics, brain-machine interfaces, and communicating and smart objects. In terms of biometrics, we work on contemporary issues to offer solutions in various fields, such as forensics, identification and authentication. As for brain-machine interfaces, we are working on thought interpretation algorithms for the purpose of assisting people with disabilities and brain activity classification algorithms to try and understand the behaviours associated with brain activity. In the field of connected objects, we are working on methodologies that can then be applied in various contexts, such as smart cities and buildings, home automation and e-health. The techniques and tools that we use in our various proposals and projects are based on artificial intelligence (neural networks, deep learning), data science (data processing, feature engineering), and metaheuristic optimisation (genetic algorithms, ant colony, simulated annealing and particle swarm). The IUSD research group works with a number of leading industrial players (e.g. RTE, EDF) and national and international research organisations (e.g. MIT, Stanford University, Northumbria University). Furthermore, the research group is involved in several funding proposals.

The research group PASTIS focuses on the modelling and understanding of human cognitive activity in interaction with digital systems with a special interest in programming, typically through simulations and the use of practical and usable tools. The members of PASTIS apply this common approach to the areas of AI, language, privacy, images and rendering, gaming, non-classical logic and belief modelling, and symbolic dynamics. The name, “PASTIS”, is also an acronym — Programming, Artificial intelligence, Security, Texts, Images, Simulations — that reflects the diversity of the different areas of computer science researched by the group.

Research areas:

• Smart Data Spaces (EID)
• Ubiquitous Computing and Data Science (IUSD)
• Programming, Artificial intelligence, Security, Texts, Images, Simulations (PASTIS)