Automating Dynamic Consent Decisions for the Processing of Social Media Data in Health Research

Area of Science:

• Digital Health
• Bioethics
• Machine Learning Applications

Background:

• Social media platforms are increasingly utilized as data sources for health research.
• Traditional informed consent methods for social media data are often impractical, burdensome, or overly broad.
• Ethical considerations regarding data privacy and participant consent are paramount in digital health research.

Purpose of the Study:

• To explore the potential of machine learning for automating granular consent decisions in health research using social media data.
• To investigate the accuracy of predicting appropriate data flow for health-related social media information.
• To identify and discuss the ethical and technical implications of implementing automated consent techniques.

Main Methods:

• Conducted an exploratory user study with 67 participants.
• Employed machine learning algorithms to predict and automate granular consent decisions for health-related social media data.
• Assessed the accuracy of data flow prediction and the minimization of undesired data leaks.

Main Results:

• Machine learning models demonstrated reasonable accuracy in predicting the appropriate flow of health-related social media data.
• The proposed technique effectively minimized undesired data leaks, enhancing data privacy.
• The study identified significant real-world implications for the practical application of automated consent.

Conclusions:

• Automated granular consent using machine learning offers a promising middle-ground approach for ethical social media health research.
• Further research is needed to address the complex ethical and technical challenges associated with this innovative technique.
• This approach has the potential to improve data accessibility while safeguarding participant privacy.