Data-driven trust scoring system for South African banks — helping consumers make informed decisions based on complaints, regulatory sanctions, and fraud resolution
# 🏦 SA Bank Trust Score — Consumer Protection Intelligence
> A data-driven trust scoring system that helps South African consumers choose a bank based on verified evidence, not marketing.
---
## 🔗 Demo Link
👉 View Live Streamlit Dashboard
---
## 📋 Table of Contents
- Business Understanding
- Technologies
- Setup
- Approach
- Status
- Credits
---
## 💼 Business Understanding
South African consumers are increasingly exposed to banking failures — unresolved complaints, regulatory violations, fraud, and poor service. Yet most consumers choose a bank based on advertising, not evidence.
This project changes that.
Using verified data from official regulatory bodies, this system builds a transparent **Trust Score** for each of South Africa's major retail banks — giving consumers, journalists, and researchers a data-driven basis for one of the most important financial decisions they will make.
Banks are scored across four verified dimensions:
| Dimension | What It Measures |
|---|---|
| 📋 **Complaint Resolution** | How quickly banks resolve complaints before Ombudsman escalation |
| ⚖️ **Consumer Favour Rate** | % of formal cases decided in the consumer's favour |
| 🔒 **Regulatory Record** | Total financial penalties from the SARB Prudential Authority |
| 💬 **Consumer Sentiment** | Social media net sentiment + consumer satisfaction surveys |
Each bank receives a final Trust Score out of 10:
| Score | Rating |
|---|---|
| 🟢 **7–10** | High Trust — strong track record across all dimensions |
| 🟡 **4–7** | Medium Trust — mixed performance |
| 🔴 **0–4** | Low Trust — significant concerns across multiple dimensions |
---
## 🛠️ Technologies
**Languages & Environment**
**Data & Analysis**
**Deployment**
| Category | Tools |
|---|---|
| **Data & Analysis** | Python, Pandas, NumPy, Matplotlib, Seaborn |
| **Notebook** | Jupyter |
| **Dashboard** | Streamlit Cloud |
---
## ⚙️ Setup
```bash
# Clone the repository
git clone
github.com …