Health Behaviour Informatics

We are living in a digital world, and increasingly the boundaries between humans and their digital environment is blurring. People can be equipped with a myriad of devices that automatically measure their physical state and often in a highly nonintrusive fashion. For example:

  • a watch that measures heart rate without the use of a chest band
  • a sensor captures various conditions of the skin, effectively measuring the level of stress that a person is currently experiencing
  • a small device with a built-in accelerometer measuring activity, and being able to distinguish walking, running, biking, and stair climbing
  • on-body GPS-enabled devices track the locations of a person providing further insight in a person’s whereabouts
  • special smartphone apps based on psychological theories assist in capturing and analyzing the moods of patients with depression-related disorders
  • proximity sensors invisibly embedded in watches, smartphones, or other on-body devices, measure local density in crowds, and combined with other sensory input, can measure a person’s “social” space.
Being able to automatically measure a person’s state allows us to automatically analyze that state and, in principle, take measures to influence that state.


A big challenge is to addresses the full cycle of automatically measuring, analysing, and acting upon the physiological and psychological state of individuals or groups of people. Emphasis lies on promoting adherence: enabling people to willingly follow a therapy or stick to an agreed set of health behaviours. Typical examples for adherence range from the ensuring that patients adhere to their therapy (e.g. medication adherence) to the case of staying active to prevent obesity (physical activity promotion). To support adherence, the basic human needs of autonomy, competence and belonging need to be incorporated into the design of health behaviour support systems.

Computational health behaviour support systems address the problems of adherence by the development of computational models and systems by which adherence is supported in an automated fashion. It explores the boundaries of large-scale, distributed techno-social systems that sense and influence behaviour of people in their social context. From a scientific perspective it faces a number of challenges:

  • What are effective methods for obtaining and deriving the factors that determine (some target) behaviour?
  • How to capture human behaviour in computational models?
  • How to predict behaviour from (partial and incomplete) data with sufficient accuracy?
  • How to use the social network of individuals to understand and measure the social health of these individuals?
  • How to deliver timely feedback to promote adherence?
  • How to verify the effectiveness of computational adherence systems on influencing people’s behaviour?

The sheer volume of data and the frequency of updates of this data, as well as as the diversity of their origins, requires research in special computational methods and techniques for data extraction (from sensing devices), data processing (massive online and real-time solutions), and actuation (effective feedback solutions). Matters are complicated by the fact that most observed behavior will emerge from unknown or poorly understood sources, requiring that most, if not all solutions, are highly adaptive. In this sense, research into computational adherence lies at the heart of the next generation of highly adaptive socio-technical systems.