JMIR Publications

ABSTRACT

Background:

Although physical activity (PA) has positive effects on health and well-being, physical inactivity is a worldwide problem. Digital health interventions are shown to be effective in promoting PA. Personalizing persuasive strategies improves intervention success and can be performed with machine learning (ML). For PA, several studies addressed personalized persuasive strategies without ML, while others included personalization with ML without focusing on persuasive strategies. An overview of studies discussing ML to personalize persuasive strategies in PA promoting interventions and corresponding recommendations could be helpful for such interventions to be designed in the future, but is missing yet.

Objective:

First, we aim to provide an overview of implemented ML techniques to personalize persuasive strategies in digital interventions promoting PA. Moreover, we aim to present a first step towards a recommendation framework as guideline for applying ML techniques in this field.

Methods:

A scoping review was conducted based on the framework of Arksey and O’Malley and the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Scoping Reviews) criteria. Scopus, Web of Science, and Pubmed were searched for studies that included ML to personalize persuasive interventions promoting PA. Papers were screened with ASReview software. From the included papers, categorized by the research project they belonged to, we extracted data regarding general study information, target group, PA intervention, implemented technology, and study details. Based on the analysis of this data, a starting point for a recommendation framework was given.

Results:

40 papers belonging to 27 different projects were included. These papers could be categorized in four groups based on their dimension of personalization. Then, for each dimension one or two persuasive strategy categories were found together with a type of ML. The overview resulted in a starting point for a recommendation framework consisting of three levels: 1) dimension of personalization; 2) persuasive strategy; and 3) type of ML. When personalizing the timing of messages, we recommend to implement RL to personalize the timing of reminders and SL to personalize the timing of feedback, monitoring, and goal setting messages. Regarding the content of messages, we recommend to implement SL to personalize PA suggestions and feedback or educational messages. For personalizing PA suggestions, SL can either be implemented alone or combined with RS. Last, RL is recommended to personalize the type of feedback messages.

Conclusions:

The overview of all implemented persuasive strategies and their corresponding ML methods is insightful for this interdisciplinary field. Moreover, it led to a starting point for a recommendation framework guiding the design and development of personalized persuasive strategies to promote PA. Based on future papers, the recommendation framework might be expanded with additional layers to specify ML methods or additional dimensions of personalization and persuasive strategies.