DataCure

DataCure is a diverse, cross-cultural research team. United by a passion for mental health and data-driven solutions, we combine expertise in data science, natural language processing, and social research to tackle real-world health challenges. Our goal is to create insights that lead to more empathetic, ethical, and effective digital mental health tools.

---

👥 Meet the team on Github

Huda Alamassi, Malak Battatt, Sadam Husen Ali, Chrismy Augustin, F. Ismail SAHIN, Aziz Azizi Ayham Hasan.

---

🔍 Project Focus

• Track: Collaborative Data Science Project (CDSP)
• Team Name: DataCure 🧬
• Title: The Public Failure Analysis: Identifying Conversational Failures in Mental Health Chatbots
• Domain: Digital Mental Health & Emotional Support Technologies
• Timeline: May to August 2025
• Current Status: 🛠️ In Progress (Milestone 5)

---

🔍 Explore Our Project

📄 To read the full background and detailed version, see Milestone 1: Problem Identification Milestone 2: Data Collection

🔹Problem Statement

📄 To read the full Problem Statement, see Problem Statement

• Chatbot vs Human vs Hybrid Support in Mental Health Apps

Mental health apps are bridging global care gaps, especially where stigma, cost, or shortages limit traditional access. Three main support models exist:

• 🤖 Chatbots: 24/7, private, and scalable, but lack emotional depth.
• 🧑‍⚕️ Humans: Empathetic and nuanced, yet costly and less available.
• 🔁 Hybrids: Blend both, offering balance but may confuse users.

Each model has pros and cons, with success shaped by user context and needs.

• Why It Matters

  • Mental health affects 1 in 8 people globally, but access remains unequal.

  • Chatbots offer 24/7 low-cost support, but may lack empathy.

  • Humans provide depth, but are less scalable.

  • Hybrids aim to combine both—but how well do they work?

• Goals

  • Systematically identify and categorize the most common user-reported failures of mental health chatbots.
  • Quantify the prevalence of these failure themes using public app store and forum data.
  • Compare failure themes between purely conversational AIs and a baseline wellness app to isolate unique "conversational" failures.

• In Summary

To explores how mental health support—via chat-bots, humans, or hybrid systems—is experienced across cultures. Our aim: to inform more empathetic, effective, and inclusive digital care.

🔹Summary of Problem Understanding

📄 To read the full Summary of Problem Understanding, see Summary of Problem Understanding

We used divergent thinking to explore broad problem areas, then applied
convergent thinking to narrow our focus using feasibility and impact criteria. We chose to explore chatbot vs human support in mental health apps.

• Idea Evaluation Sheet

  All proposed project ideas and their comparison scores are documented in this spreadsheet This helped us select the most impactful and feasible topic.

Iceberg Model Overview

• Event: Users interact with chatbots—some feel helped, others feel unheard.
• Pattern: Mental health apps are growing; many now rely on chatbots.
• Structure: Cost and scale drive chatbot use; regulation is limited.
• Mental Model: Belief in AI as a fix for care gaps and stigma sustains the system.

🔹Research Question

📄 To read the full Research Questions, see Research Question

Our refined research question, which is achievable through public data analysis, is:

What are the most prevalent themes of user-reported conversational failure in leading mental health chatbots, and what do these themes reveal about the gap between user expectations for emotional support and current algorithmic capabilities?

This helps us focus our research on a specific, measurable aspect of the problem.

---

📁 Directory Structure

/
├── README.md                   - Project overview and main instructions
├── guide.md                    - Detailed guide on using this template
├── /collaboration/             - Team norms, strategies, and retrospectives
├── /notes/                     - Shared resources and learning materials
├── /0_domain_study/            - Domain research and background
├── /1_datasets/                - Raw and processed datasets
├── /2_data_preparation/        - Scripts for cleaning and processing data
├── /3_data_exploration/        - Scripts for initial data understanding
├── /4_data_analysis/           - Scripts for in-depth analysis
├── /5_communication_strategy/  - Materials for communicating findings
└── /6_final_presentation/      - Final presentation materials

🤝 Our Working Agreements

• 🔹 Team Norms
• 🔹 Communication Plan
• 🔹 Constraints
• 🔹 Learning Goals

---

⏳ Milestones & Timeline

Milestone	Description	Progress	Target Date
0	Team Setup & Collaboration	✅ Complete	June 2
1	Define Research Question	✅ Complete	June 16
2	Data Collection	✅ Complete	June 30
3	Analysis & Modeling	✅ Complete	July 21
4	Communication Strategy	✅ Complete	August 11
5	Final Presentation	⏳ Upcoming	August 25

---

🔍 Explore Milestones

• 🔹 Milestone 0: Cross-Cultural Collaboration
• 🔹 Milestone 1: Problem Identification
• 🔹 Milestone 2: Data Collection
• 🔹 Milestone 3: Data Analysis
• 🔹 Milestone 4: Communicating Results
• 🔹 Milestone 5: Final Presentation

---

➕ Contributing

We’re excited to collaborate! To get involved, please check out our CONTRIBUTING.md for guidelines.

---

🔑 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

“There is no health without mental health.” – World Health Organization