Prevalence Overshadows Concerns? Understanding Chinese Users’ Privacy Awareness and Expectations Towards LLM-based Healthcare Consultation

This repository contains the code and redacted datasets for our IEEE S&P 2025 paper: Prevalence Overshadows Concerns? Understanding Chinese Users’ Privacy Awareness and Expectations Towards LLM-based Healthcare Consultation

---

REDACTED Datasets

This repository provides datasets where sensitive information has been removed to address ethical concerns. Specifically:

• Demographics & Background sections have been removed.
• Responses containing free-text answers have been marked as REDACTED.

All text or statistical features that could directly or indirectly identify participants have been deleted.

Data Analysis

To explore the data analysis details, please refer to the following Jupyter Notebooks:

• DATA_Analysis_Part I - Awareness During Consultation (RQ1).ipynb
• DATA_Analysis_Part II - Privacy Expectations (RQ2).ipynb
• DATA_Analysis_Part III - Attitudes and Previous Experiences (RQ3).ipynb

The scripts provided in these notebooks have been processed to handle any sensitive information. Note that some datasets (e.g., DataforCI_ShortColumnName.xlsx and 00Codebook) have been hidden from the release to maintain privacy.

LLM-based Healthcare Chatbot

The deployed system can be accessed at:

• Chinese version
• English version

The HealthcareChatbot directory contains the local version of the Healthcare Chatbot developed using the Django framework. This chatbot provides a responsive user interface and a survey questionnaire interface across multiple platforms. The system includes the informed consent form, the healthcare chatbot interface, and the embedded questionnaire. Note: Some Quality Control methods, such as restricting repeated participation in the survey, have been removed from this release.

Running the Project Locally

If you have experience with running Django projects, follow these steps to run the Healthcare Chatbot locally:

1. Review settings.py and views.py: Update or replace necessary fields, particularly the DATABASES and openai configuration.

2. Install dependencies: pip install -r requirements.txt or pip install -r simple-requirements.txt

   If any packages fail to install, please manually install them.

3. Perform database migrations:

   python manage.py makemigrations
   python manage.py migrate

4. Run the development server:

   python manage.py runserver

5. Access the system: Open a web browser and go to:

   • http://127.0.0.1:8000/LLMHPri/encindexzh/ for the Chinese version.
   • http://127.0.0.1:8000/LLMHPri/encindexen/ for the English version.

---

Citation

Please consider citing the following paper if you found our work useful.

Zhihuang Liu, Ling Hu, Tongqing Zhou, Yonghao Tang, Zhiping Cai. Prevalence Overshadows Concerns? Understanding Chinese Users’ Privacy Awareness and Expectations Towards LLM-based Healthcare Consultation, in 2025 IEEE Symposium on Security and Privacy (SP), San Francisco, CA, USA, 2025, pp. 2493-2511, doi: 10.1109/SP61157.2025.00092.