A systematic review on eHealth technology personalization approaches.

Area of Science:

• Digital Health
• Human-Computer Interaction
• Computational Methods

Background:

• Personalization of eHealth technologies is common but not well understood.
• A comprehensive understanding of personalization approaches and their computational underpinnings is lacking.

Purpose of the Study:

• To systematically review and cluster personalization approaches in eHealth.
• To identify computational methods used for user segmentation and technology adaptation.
• To reveal gaps in current personalization research.

Main Methods:

• Systematic literature review of 412 reports on eHealth personalization.
• Clustering of personalization approaches based on user segmentation variables and eHealth adaptations.
• Identification and categorization of computational methods employed.

Main Results:

• Identified 13 distinct clusters of personalization approaches (e.g., behavior + channeling, environment + recommendations).
• Documented 10 computational methods used to match user segments with technology adaptations (e.g., classification, reinforcement learning).
• Found limited exploration of technology-related variables and user interaction reminders.

Conclusions:

• Current eHealth personalization relies heavily on single variable types, indicating a need for more sophisticated segmentation.
• Future research should integrate technology-specific features to develop more individualized personalization strategies.
• Further investigation into user interaction reminders and diverse variable types is recommended for enhanced eHealth personalization.