File size: 1,733 Bytes
e369950 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
Fake Profile Detection
# 🚨 Fake Instagram Profile Detection using Machine Learning
This project is a real-time Instagram profile analyzer that predicts whether a given profile is **fake** or **real** using machine learning. It uses profile metrics like follower count, following count, post count, and verification status to make predictions.
---
## 📌 How It Works
- You enter an Instagram **username**.
- The application uses the **Apify API** to fetch public profile data.
- It extracts key features such as:
- Number of followers
- Number of followings
- Number of posts
- Is the account private?
- Is the account verified?
- These features are passed into a pre-trained **machine learning model** (`classifier.pkl`) to predict whether the profile is real or fake.
---
## 🛠 Technologies Used
- **Python**
- **Streamlit** – for building the web app
- **Joblib** – for loading the ML model
- **Apify API** – to scrape Instagram data
- **Scikit-learn** – for training the ML model
- **Pandas, NumPy** – for data manipulation
---
## 🧠 ML Model
The model is trained using a labeled dataset containing Instagram profile attributes. The classification is binary:
- `0` → Likely Fake
- `1` → Likely Real
The training includes feature normalization and multiple algorithm trials like Logistic Regression, Decision Trees, and Random Forests. The final deployed model is chosen based on accuracy and generalization.
---
## 🖥️ Project UI
- The app is built with **Streamlit** for a clean and interactive interface.
- Users simply input a **username** and click **Predict**.
- Output shows the profile’s stats and the prediction result with appropriate messaging (Success/Error).
---
|