Negotiation Behaviors of Owners and Bystanders over Data Practices of Smart Home Devices

Document Title

Exploring the Negotiation Behaviors of Owners and Bystanders over Data Practices of Smart Home Devices

Document Information

• Subject Area: Smart home data privacy management and negotiation behaviors
• Keywords: Smart home devices, data practices, privacy protection, negotiation behaviors, digital agents

Research Background and Issues

• Problems and Challenges: Research on smart home privacy protection typically focuses on device owners, with less attention paid to the privacy risks faced by bystanders (e.g., visitors or domestic helpers) who are temporarily present in smart home environments. These bystanders often lack understanding of the data practices of smart home devices and may face risks of privacy violations.
• Significance: With the increasing prevalence of smart home devices, there is an urgent need for mechanisms that ensure security while respecting others' privacy. Exploring how to resolve conflicts between owners and bystanders through negotiation is a critical direction in the field of smart home privacy protection.
• Motivation and Related Work: Some studies suggest enhancing privacy protection by increasing transparency in data practices or empowering bystanders. However, there is currently a lack of in-depth research on how bystanders negotiate data collection, storage, and sharing practices.

Solution

• Methods and Design:
  • Proposed an experimental design for negotiation behaviors based on digital agents. Participants were assigned roles as device owners or bystanders and interacted with simulated digital agent characters to explore negotiation behaviors.
  • Designed three external factors (agent type, relationship type, negotiator type) and two internal factors (smart home device type and data practice type) to simulate negotiation processes in different scenarios.
• Innovations:
  • Utilized digital agents to simulate negotiation counterparts for participants. Two types of agents—strict and collaborative—were designed to flexibly model different negotiation behaviors. This design provides a controlled experimental environment for in-depth research.
  • Randomized experimental scenarios in the design to control for external confounding variables.
• Implementation Steps and Techniques:
  • Implemented negotiation agents through a web application, recording participants' full interactions with the agents (preference changes, number of rounds, agreements reached, etc.).
  • Adopted the Rubinstein Bargaining model framework, allowing both parties to engage in multi-round negotiations.

Research Findings

• Specific Results:
  • Significant differences were observed in the negotiation behaviors of owners and bystanders: owners were more inclined to maintain the utility of the devices, while bystanders were more willing to compromise to protect privacy.
  • The number of negotiation rounds indicated that the process was not overly burdensome, with most agreements reached within 1-3 rounds.
  • The type of data practice significantly influenced negotiation outcomes; for instance, data sharing practices were more easily agreed upon compared to data collection and storage.
• Advantages over Existing Solutions:
  • Provided a comprehensive research framework for negotiation behaviors, offering new insights into smart home data privacy management.
  • The findings directly support the development of smart home data negotiation features that empower both bystanders and owners.
• Experimental or Evaluation Results:
  • The average accuracy of predicting agreement outcomes was 70%, with a maximum of 78.2% (for data collection practices of smart cameras).
  • Based on participants' behaviors and responses, several key factors influencing negotiation behaviors were identified, such as the presence of negotiation history and device type.
• Limitations and Future Directions:
  • The sample primarily consisted of technologically proficient AMT workers, which may not fully represent the general population.
  • Future research could expand scenario designs to include complex interactions involving multiple owners and bystanders.
  • Explore the feasibility of integrating digital agents into actual smart home devices for automated negotiation functionalities.