Motivating Physical Activity through a combination of Activity Tracking, Context-Awareness, Recommender Systems and User Centred Design approaches

Contact person: Professor George A. Papadopoulos

Alexandros Yeratziotis
Christos Mettouris
Achilleas Achilleas

Project Description

Lack of and decline in physical activity amongst all age groups has led to an increase of health problems, both physical and mental. One reason that contributes to inactivity is motivation absence, which has been at the centre of research and developments in the area.

Activity tracking devices are available in the retail market attempting to fulfil this gap of motivation absence in order to increase users' physical activities and to also assess the quality of their activity patterns (e.g. good exercise, quality sleep). With smartphones, we are now able collect sensor data, analyze it and detect patterns and causalities from it. This enables understanding users' behaviour in an effort to show meaningful recommendations at the right time and in the correct format without the least user annoyance.

Context-aware services incorporate information about mobile users such as their current location, as well as information from the environment such as the weather, so that relevant context-aware and personalized services can be provided to the users (e.g. real-time traffic update). Temperature, time and location are examples of real-world context parameters used. The users themselves can manually update some of this information if needed, however, in more recent scientific works, contextual information is being updated dynamically, based on device-device and sensors-device communications.

Achieving long-term motivation is difficult; especially when the aim is to convince inactive people to be more active, even if it is for their own good. Providing useful services to users can contribute to this effort. These services can be activity related and can be recommended to users with a purpose of engaging them into a physical activity. This entails creating and evaluating concepts for the recommended services, and ensuring that the technology used to present these is appealing, useable and enjoyable to use for users of all age groups.

We believe that by considering several topics within ICT this can be achieved more successfully. The topics of activity tracking, context-awareness, recommender systems and user-centred design will be considered in this study for this specific purpose. The outcome of this multidisciplinary approach will be a mobile application that offers users personalized, context-aware recommendations, tailored to their age and interests, and which intends to engage them in different forms of physical activity. Recommendations need to be provided to users at the right time and by considering their context.

Considering activity tracking, context-awareness, recommender systems and user-centred design in tandem, can lead to increased motivation for physical activity in a more intelligent and unobtrusive manner. They can provide the "nudge" towards a good lifestyle. In turn, adoption rates of the mobile application can be increased and extended periods of use be observed in comparison to other existing technologies, which have not considered all of these topics together for the purpose of motivating users for physical activity. In conclusion, the research will:

  • Investigate existing activity tracking methodologies and technologies and determine the most appropriate ones that will cater for a diversity of users.
  • Investigate recommendation algorithms and determine the most appropriate ones that will be integrated into the mobile app.
  • Investigate context-sensing and context-awareness methodologies to determine which/how will be applied/can be used with smartphones and activity tracking and what information will be collected/sensed.
  • Create and validate concepts for services that can be recommended to users and which will require them to engage in physical activity. Methods from HCI can be used for this purpose. This can be thought of as a form of reverse engineering, since ideas on the concepts and design of services follows the already existing technologies (algorithms for activity tracking and recommendations, context-aware systems).
  • Apply user-centred design approaches to design an application that is usable and provides positive user experiences to users of all ages. This entails understanding what the most valuable information for users is, how it should be presented to them and when it should be presented to them.

Other Postdoc topics on Software Engineering for Mobile Applications for the ERCIM Fellowship programme. Similar topics for PhD positions.

In the last years, context-awareness and mobile computing has seen popular adoption connecting also with Service Oriented Computing (SOC) and the Internet of Things (IoT). Relevant hardware equipment can be found in smartphones, smartwatches, sensors and microcontroller boards, such as Arduino, Raspberry pi and the recent CHIP. Context-aware mobile computing in this framework with IoT and smart Cyber Systems will change the way we interact with objects and services, but includes many challenges. For instance, access to context requires access to sensitive information concerning users and their surroundings. Deep understanding of data sharing consequences by mobile platform users is limited (e.g. Android OS, iOS); users usually accept all requirements (e.g., when installing an application) without understanding well if there are potential violations on the privacy sphere related to the application use.

To address the above there are postdoc (and PhD) opportunities at the Software and Internet Technologies (SEIT) laboratory of the Department of Computer Science.If this is of interest to you, you can consider more working on topic like the following:

  • Extraction of context profiles from different sources: This topic involves the investigation of techniques and methods that can lead to the extraction of context-relevant data from user models/profiles in social media and other sources.
  • Privacy by Design methods in application development for mobile platforms: This topic involves the design a specification model to capture privacy requirements for privacy-aware native mobile platform application development with a focus on Android and the enrichment of current solutions that respect privacy options in native mobile platform environments through IDE-based plugins (e.g., Eclipse).
  • Privacy design patterns recommendations: This topic involves building privacy design patterns at platform independent level, tailor them to mobile platforms, linking them with corresponding code excerpts and refining them using contextual information, and building and recommending privacy protecting coding best practices.

There are also topics available on Software reuse in Software Engineering:

  • Mining software repositories and developer expertise: The developer social coding and Q&A websites are sources of valuable information that can be combined to draw different conclusions on users' activity and expertise. This topic involves the intelligent combination of information from various sources in order to draw meaningful conclusions on developers' knowledge.

For any further information, please contact Georgia Kapitsaki