Search results

Machine Learning Methods to Personalize Persuasive Strategies in mHealth Interventions That Promote Physical Activity

Background: Although physical activity (PA) has positive effects on health and well-being, physical inactivity is a worldwide problem. Mobile health interventions have been shown to be effective in promoting PA. Personalizing persuasive strategies improves intervention success and can be conducted using machine learning (ML). For PA, several studies have addressed personalized persuasive strategies without ML, whereas others have included personalization using ML without focusing on persuasive strategies. An overview of studies discussing ML to personalize persuasive strategies in PA-promoting interventions and corresponding categorizations could be helpful for such interventions to be designed in the future but is still missing. Objective: First, we aimed to provide an overview of implemented ML techniques to personalize persuasive strategies in mobile health interventions promoting PA. Moreover, we aimed to present a categorization overview as a starting point for applying ML techniques in this field. Methods: A scoping review was conducted based on the framework by Arksey and O’Malley and the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) criteria. Scopus, Web of Science, and PubMed were searched for studies that included ML to personalize persuasive strategies in interventions promoting PA. Papers were screened using the 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. On the basis of the analysis of these data, a categorization overview was given. Results: In total, 40 papers belonging to 27 different projects were included. These papers could be categorized in 4 groups based on their dimension of personalization. Then, for each dimension, 1 or 2 persuasive strategy categories were found together with a type of ML. The overview resulted in a categorization consisting of 3 levels: dimension of personalization, persuasive strategy, and type of ML. When personalizing the timing of the messages, most projects implemented reinforcement learning to personalize the timing of reminders and supervised learning (SL) to personalize the timing of feedback, monitoring, and goal-setting messages. Regarding the content of the messages, most projects implemented SL to personalize PA suggestions and feedback or educational messages. For personalizing PA suggestions, SL can be implemented either alone or combined with a recommender system. Finally, reinforcement learning was mostly used 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 categorization overview that provides insights into the design and development of personalized persuasive strategies to promote PA. In future papers, the categorization overview might be expanded with additional layers to specify ML methods or additional dimensions of personalization and persuasive strategies.

Digital health interventions to improve eating behaviour of people with a lower socioeconomic position: a scoping review of behaviour change techniques

Specific approaches are needed to reach and support people with a lower socioeconomic position (SEP) to achieve healthier eating behaviours. There is a growing body of evidence suggesting that digital health tools exhibit potential to address these needs because of its specific features that enable application of various behaviour change techniques (BCTs). The aim of this scoping review is to identify the BCTs that are used in diet-related digital interventions targeted at people with a low SEP, and which of these BCTs coincide with improved eating behaviour. The systematic search was performed in 3 databases, using terms related to e/m-health, diet quality and socioeconomic position. A total of 17 full text papers were included. The average number of BCTs per intervention was 6.9 (ranged 3–15). BCTs from the cluster ‘Goals and planning’ were applied most often (25x), followed by the clusters ‘Shaping knowledge’ (18x) and ‘Natural consequences’ (18x). Other frequently applied BCT clusters were ‘Feedback and monitoring’ (15x) and ‘Comparison of behaviour’ (13x). Whereas some BCTs were frequently applied, such as goal setting, others were rarely used, such as social support. Most studies (n = 13) observed a positive effect of the intervention on eating behaviour (e.g. having breakfast) in the low SEP group, but this was not clearly associated with the number or type of applied BCTs. In conclusion, more intervention studies focused on people with a low SEP are needed to draw firm conclusions as to which BCTs are effective in improving their diet quality. Also, further research should investigate combinations of BCTs, the intervention design and context, and the use of multicomponent approaches. We encourage intervention developers and researchers to describe interventions more thoroughly, following the systematics of a behaviour change taxonomy, and to select BCTs knowingly.

The identification of synthetic drug laboratories based on the detection of precursors

.

Mobile multimedia technology, a process for consumer interaction and brand penetration

Dynamic and probabilistic safety zones for autonomous mobile robots operating near humans

The inefficiency of maintaining static and long-lasting safety zones in environments where actual risks are limited is likely to increase in the coming decades, as autonomous systems become more common and human workers fewer in numbers. Nevertheless, an uncompromising approach to safety remains paramount, requiring the introduction of novel methods that are simultaneously more flexible and capable of delivering the same level of protection against potentially hazardous situations. We present such a method to create dynamic safety zones, the boundaries of which can be redrawn in real-time, taking into account explicit positioning data when available and using conservative extrapolation from last known location when information is missing or unreliable. Simulation and statistical methods were used to investigate performance gains compared to static safety zones. The use of a more advanced probabilistic framework to further improve flexibility is also discussed, although its implementation would not offer the same level of protection and is currently not recommended.

Reconstructing with trace information

Currently, promising new tools are under development that will enable crime scene investigators to analyze fingerprints or DNA-traces at the crime scene. While these technologies could help to find a perpetrator early in the investigation, they may also strengthen confirmation bias when an incorrect scenario directs the investigation this early. In this study, 40 experienced Crime scene investigators (CSIs) investigated a mock crime scene to study the influence of rapid identification technologies on the investigation. This initial study shows that receiving identification information during the investigation results in more accurate scenarios. CSIs in general are not as much reconstructing the event that took place, but rather have a “who done it routine.” Their focus is on finding perpetrator traces with the risk of missing important information at the start of the investigation. Furthermore, identification information was mostly integrated in their final scenarios when the results of the analysis matched their expectations. CSIs have the tendency to look for confirmation, but the technology has no influence on this tendency. CSIs should be made aware of the risks of this strategy as important offender information could be missed or innocent people could be wrongfully accused.