"I know what you did last semester": Understanding Privacy Expectations and Preferences in the Smart Campus

Title of the Paper

"I know what you did last semester": Understanding Privacy Expectations and Preferences in the Smart Campus

Paper Information

• Topic Area: Privacy protection and user preferences in smart campuses
• Keywords: Privacy, sensor data collection, smart campus, experience sampling method, control rules, sensing technologies, academic environment, data sharing, notification frequency, privacy expectations

Research Background and Issues

• Identified Issues or Challenges:

  • The widespread deployment of sensing technologies in smart campuses raises privacy concerns, such as insufficient transparency in data collection and the inability of users to personalize data control.
  • Sensor-collected data in academic environments may involve sensitive human behavior information, yet users have limited understanding of these technologies and their privacy implications.
  • Previous studies primarily focused on developing privacy protection technologies without thoroughly investigating users' privacy expectations and preferences in smart campuses.

• Significance of the Research:

  • Smart campuses have the potential to enhance sustainability and interconnected environments, but they also need to address the complex relationships between students, faculty, and sensing technologies.
  • Understanding the impact of indoor locations, data access roles, and data usage on users' privacy perceptions can help optimize sensor deployment strategies in smart campuses.

• Motivation and Related Work:

  • Existing research shows that sensor deployment locations and data usage purposes significantly affect privacy perceptions in other domains (e.g., homes and public spaces). However, the unique privacy challenges posed by subspaces in academic environments (e.g., study areas, recreational spaces) remain underexplored.
  • Common privacy concerns in online services and IoT environments (e.g., data storage duration and access permissions) may similarly influence user experiences in smart campuses.

Proposed Solution

• Proposed Methods or Solutions:

  • Conducted a 14-day investigation into privacy perceptions and preferences in smart campuses using the Experience Sampling Method (ESM) to deeply explore users' attitudes toward indoor sensing technology applications.
  • Leveraged a mobile application to present sensor scenarios (including sensor types, purposes, locations, etc.) based on users' actual locations, collecting real-time data on users' privacy perceptions and control rule preferences.

• Innovative Contributions:

  • Quantitative analysis identified key factors influencing users' privacy attitudes in academic environments, including indoor location types and data access roles.
  • Proposed sensor control rule-setting schemes based on specific scenarios, highlighting privacy protection needs in "group areas" and similar spaces.
  • Unveiled the unique characteristics of rule design in academic spaces, reflecting the influence of relationships and power dynamics between students and faculty on privacy expectations.

• Implementation Steps and Key Technologies:

  1. Participants completed demographic questionnaires and basic privacy perception tests.
  2. A smartphone application pushed survey questions in real-time based on users' locations, inquiring about privacy perceptions related to indoor location types (e.g., comfort level, notification preferences, allow/deny data collection).
  3. Additional in-depth surveys were conducted in the evening to explore participants' reactions to daytime scenarios and preferences for sensor control rules.
  4. Data was stored in the cloud (AWS DynamoDB) and locally on participants' smartphones for backup.

Research Findings

• Specific Findings:

  • Indoor location types (e.g., study areas, private spaces) and data access roles (e.g., faculty) significantly influenced users' privacy perceptions, including comfort levels, notification preferences, and decisions to allow/deny data collection.
  • Users were particularly concerned about sensor deployment in "study areas" and "private spaces," often preferring to deny data collection in these locations.
  • Specific groups in academic environments (e.g., faculty) raised notable privacy concerns regarding data access permissions.

• Comparison with Existing Solutions and Advantages:

  • Unlike other studies that primarily focus on sensor types and data usage, this research emphasized the importance of indoor location types, directly addressing the practical needs of academic environments.
  • Provided actionable recommendations for sensor control rule design, helping smart campus designers better balance privacy protection and sensor deployment efficiency.

• Experimental or Evaluation Results:

  • GLMM (Generalized Linear Mixed Model) analysis revealed that indoor location types and data access roles were the primary factors influencing privacy perceptions, while sensor types and data usage purposes were less significant.
  • Specific statistical results indicated the highest rejection rate for data collection in private spaces and the lowest notification preference in service areas.
  • Participants were more inclined to set data control rules for "group areas," focusing on data usage goals, notification frequency, storage duration, and access permissions.

• Limitations and Future Directions:

  • Limitations:
    • Participants were primarily students, which may not fully represent the privacy perceptions of faculty and other user groups.
    • The sample was drawn from a single university in a mid-sized U.S. city, limiting applicability to broader cultural contexts or educational institutions.
    • The two-week study duration was relatively short, failing to capture long-term changes in privacy awareness.
  • Future Directions:
    • Extend research to diverse cultural contexts and a broader range of academic institutions.
    • Conduct more detailed privacy preference analyses for different indoor subspaces and "other personnel" roles.
    • Design long-term privacy perception studies to observe changes in user cognition over time.

Conclusion

This study utilized the Experience Sampling Method to deeply investigate users' privacy expectations and control rule preferences regarding sensor deployment in smart campuses, highlighting the importance of indoor location types and data access permissions in academic environments. The findings provide valuable recommendations for smart campus design, aiding in the development of more user-friendly and transparent sensor monitoring policies.