External links

Research of Excellence on Digital Technologies and Wellbeing

LM2023061 [Registered results] [WWW] 2024 - 2028

The DigiWELL project responds to research challenges in key areas of the rapidly evolving domain of digital technologies aimed at promoting physical, mental and social wellbeing. Due to the biopsychosocial perspective we take in the project, the project is interdisciplinary in nature. Specifically, the project consists of six research work packages (WPs) that are firmly anchored in the field of psychology as the primary expert domain. However, other disciplines are also applied within the individual WPs, which is reflected in the composition of the implementation team, which includes experts from the fields of psychology, media studies and communication, kinanthropology, sociology, social work, mental health, public health, statistics and computer science. Research Work Package in which the ICS is collaborating, will focus on the research, development, and application of statistical models of knowledge-based analytical methods for processing information burdened with uncertainty, missingness and unstructured text in the field of longitudinal intensive, biosensory, and social science data as well as other methods of processing time series, enabling the detection and prediction of changes in behavior and wellbeing. Several statistical and non-statistical approaches will be used, including cluster analysis and machine learning methods based on deep neural networks. Data will be modeled using state-of-the-art statistical approaches, including semiparametric, dynamic, and spatial models. Both frequentist and Bayesian methods will be used, enabling the fusion of empirical and expert information, using computationally efficient approximations suitable for big data. Analyses will serve both to explore and test new structural dependencies in data motivated by psychological theories and their extensions, and to develop models suitable for future use for predictively derived interventions. These will include state-space models, which will enable the sequential estimation of latent characteristics, on which it is then possible to build formalized intervention algorithms for behavior change.

Grant registration number: CZ.02.01.01/00/22_008/0004583

This project is co-funded by the EU.