Eliciting Security & Privacy-Informed Sharing Techniques for Multi-User Augmented Reality

Title of the Paper

Eliciting Security & Privacy-Informed Sharing Techniques for Multi-User Augmented Reality

Paper Information

• Subject Area: Multi-user interaction design in Augmented Reality (AR), with a focus on security and privacy (S&P)-related content sharing techniques
• Keywords: Multi-user augmented reality, sharing techniques, privacy and security, threat modeling, user-driven elicitation study, co-design, AR content access control, scenario design, access permissions, interaction techniques

Research Background and Issues

• Problems and Challenges:

  • Current multi-user AR designs primarily focus on usability and technical feasibility, often neglecting security and privacy (S&P) requirements. As AR becomes a popular collaborative tool, particularly in education, unresolved privacy and security risks may arise, such as virtual content access control and privacy data leaks.
  • There is a lack of effective methods to guide AR designers in incorporating S&P considerations during the design process.

• Research Significance:

  • The rapid adoption of AR devices, especially in everyday life and education, has made these issues increasingly urgent.
  • Addressing these challenges will lay the foundation for designing safer and more user-accepted AR technologies.

• Research Motivation and Related Work:

  • S&P research has covered various potential threats (e.g., unauthorized access to virtual content, privacy protection in physical spaces), but it has not been sufficiently integrated with mainstream AR interaction design research.
  • Existing user-driven elicitation methods have successfully improved usability but have yet to systematically address security and privacy concerns.

Solution

• Proposed Method:

  • Combine user-driven elicitation methods with scenario design to create a process that emphasizes S&P issues.
  • Design specific scenarios (e.g., "Future of Education" scenarios) and use step-by-step guided questions to inspire experts to propose interaction techniques that balance usability, technical feasibility, and S&P requirements.
  • Employ a threat modeling-based stepwise elicitation process, introducing potential threats in specific contexts (e.g., permission management for virtual content manipulation, bystander privacy risks).

• Innovations:

  • Directly integrate security and privacy (S&P) design goals into user elicitation studies, rather than treating them as afterthoughts.
  • Design an experimental framework combining AR experts and S&P experts to explore how their collaboration affects the quality of proposed solutions.
  • Propose prototypes and improvement suggestions for interaction techniques to meet the complex design requirements of multi-user AR.

• Implementation Steps:

  1. Create multi-user AR scenarios and define design dimensions (e.g., spatiotemporal matrices, public and private spaces, threat models).
  2. Recruit experts in AR and S&P for user elicitation studies, dividing them into AR/AR and AR/SP pairing conditions.
  3. Conduct interaction design tasks with step-by-step prompts, followed by critical evaluations of the proposals using threat models.
  4. Record and analyze the interaction techniques, evaluating their balance between usability, feasibility, and S&P goals.

Research Outcomes

• Specific Results:

  • A total of 73 interaction proposals for sharing virtual content and 65 access control enhancement techniques were developed.
  • The proposals included both traditional interaction methods (e.g., virtual menus) and innovative methods (e.g., sharing based on physical markers).
  • Both experimental groups (AR/AR and AR/SP) proposed similar types of access control techniques, but the AR/SP group prioritized S&P considerations earlier in the process.

• Advantages Over Existing Solutions:

  • Compared to traditional elicitation methods, this study systematically integrates S&P concerns into the design process.
  • Provides multimodal interaction techniques (e.g., gestures, menus, gaze) to address the complexities of multi-user AR collaboration.
  • The method is applicable to relatively secure environments like education while considering broader privacy protection strategies.

• Experimental and Evaluation Results:

  • Observations from both pairing conditions showed that the creation of multi-user AR interaction techniques could balance usability, technical feasibility, and S&P goals.
  • The involvement of S&P experts significantly promoted the prioritization of S&P design goals within the pairings without stifling the creativity of AR designers.
  • Two design strategies, "usability-first" and "S&P-first," were validated.

• Limitations and Future Directions:

  • Limitations:
    • The use of experts rather than end-users may result in findings that do not fully reflect the needs of end-users.
    • The specificity of the designed scenarios may limit their generalizability to other contexts.
    • The lack of a baseline comparison with traditional methods prevents direct performance comparisons.
  • Future Work:
    • Conduct experiments in more domains/scenarios to study the generalizability of the method.
    • Collaborate with a broader user group to explore the effects of non-expert participation.
    • Develop new scenarios or threat models to explore interaction technique design in other AR application contexts.