Topics of particular focus in ESOTICS include, but are not limited to:
One notable class of social requirements which we deal with in ESOTICS relates to transparency. Our aim is to enable an informed decision on the information to be made transparent in an enterprise and a socio-technical system. Lack of transparency may be associated with privacy and security requirements, but we also recognise other equally important reasons inherent to transparency decisions such as information overload, relevance, timing, pressure, creating bias, etc.
Another class of social requirements studied in the ESOTICS group relates to motivation requirements. These requirements are typically realised via different technologies under various guises, such as gamification, persuasive technology, entertainment computing and incentive-centred design. We study the coherence between those requirements and the technology meant to achieve them on one side and the other requirements and characteristics of an enterprise, including social and mental wellbeing at the workplace, on the other. Our argument is that these mechanisms, intended to meet motivational requirements and maximise efficiency and enjoyment, may have negative side effects which need to be captured and catered for early on at the engineering stage. This includes creating unnecessary competition, reducing quality to win the reward, encouraging clustering, etc.
Social and crowd-centric software systems adaptation
To cope with the dynamic nature of social requirements, our engineering process takes a live and continuous engineering style. Our method for keeping the requirements, design models and knowledge base updated is to open a feedback channel with the actual crowd of users and stakeholders of the system. This introduces an engineering challenge relating to how to get meaningful and trustworthy feedback, without affecting users' experiences and causing additional overhead. This is also a major focus area of our research in ESOTICS.
We have a keen interest in studying digital addiction, which we define as “a problematic usage of digital devices which is characterized by properties like being excessive, obsessive, compulsive, impulsive and hasty”. The sparse literature on the topic has mainly focused on the user’s side, e.g. for the analysis of the psychological and social reasons, symptoms and side effects. Our argument is on the need for an accountable software design and development process, which treats addiction awareness as a first class requirement. The engineering of testing, predictive and preventive methods to combat digital addiction is one of our main focus areas.
The Affective Computing and Multimodal Interaction research topic deals with the study and development of systems that can recognise, interpret and process human affective states/emotions. This is realised during human-machine or human-human interaction, using multimodal measures. The challenges rely on (a) the identification and definition of affective states / emotions of interest, (b) the processing of physiological (e.g. Galvanic Skin Response, EEG) and behavioural (e.g. speech, face, acceleration) signals, (c) the modelling of the affective states and (d) the development of signal synchronisation measures for studying social settings.
Social psychology and cyber design
The internet has fundamentally changed how we communicate as individuals and how we shape our social identities in a range of online mediums including social media, online games and augmented reality. Many of these changes have been unintentional, and demonstrate the unexpected societal consequences that can be brought about by new web technologies. We aim to better understand the relationship between social psychological processes and socio-technological systems, through consideration of the social psychological literature and how this may be influenced by factors such as anonymity and disinhibition. We also argue that the design of socio-technical systems needs to take into account not just the effects on the individual but also the groups to which that individual may belong.
Model Driven Development (MDD) is a software engineering methodology that suggests the systematic use of models as the primary means of the software development process. MDD introduces model transformations between the models at different abstraction levels and proposes the use of metamodels for specifying modeling languages formally. In MDD, models are transformed into other models in order to (semi)automatically generate the final source code.
In our research, we focus on automated software engineering to increase the productivity, reusability and quality during the software development process. For example, we use metamodeling, model transformations, formal language theory, systems theory, graph theory, conceptual modeling, and various methods to move the concepts into code. Our expertise in domain specific modeling and metamodeling provides a formal basis for model based systems engineering as well. Our research is interdisciplinary and practical which provides new tools and methods as well as language specifications.
Some recent research topics that we are interested in are as follows: Modeling for Internet of Things, Software requirements modeling, MDD for Modeling and Simulation, Formal Transparency Modeling, etc.