Exploring user-friendly data privacy support in multimodal sensor-based digital healthcare contexts

Abstract

This thesis investigates the user-oriented privacy concerns and design requirements associated with collecting personal and interpersonal data in sensor-driven healthcare services. With the increasing popularity of personal informatics research powered by sensor technology, it has become crucial to address users' privacy concerns and enable them to scrutinize and regulate the representation of their personal data in pervasive computing. The extensive gathering of sensor readings poses potential privacy risks, such as revealing participants' current activities through accelerometer data or disclosing their identity through gait recognition data. These privacy threats not only endanger the participants but also the end-users of the technology, as they offer a wealth of information about an individual, including environmental conditions, activity patterns, health records, behavioral habits, and locations. Consequently, it has become essential for individual users to understand and interpret their own data while exerting control and asserting ownership over their personal information. Therefore, this thesis explores the motivations and privacy concerns of individual users regarding sharing personal data and investigates the suitability of various design features as privacy control mechanisms. The research encompasses both large-scale and small-scale studies, involving surveys, interviews, and real-world experiments. In addition to personal informatics, the thesis introduces a novel concept called "Interpersonal Informatics and Interpersonal Data," which encompasses the reflection and monitoring of interpersonal data. This includes personal data as well as co-produced data generated from interactions among multiple users (e.g., tracking physical activity together) or the utilization of digital healthcare services (e.g., a game-based family exercise application) within the context of multimodal sensor-based healthcare services. By considering this new concept, we aim to explore privacy concerns and attitudes related to both personal and interpersonal data, as well as design studies that support users in actively managing their own data. By examining both personal and interpersonal dimensions in the collection of multimodal sensor data, we discuss design strategies that can enhance user data literacy and empower users to have more control over data management and proactive privacy decisions. This thesis demonstrates the feasibility of user-friendly privacy support across various domains of sensor-driven research and services.