Bot-Detection-Captcha

Human vs Bot Detection Based on Captcha Movement Variability

This project aims to develop a robust system for detecting human and bot interactions based on the variability in movement patterns during a captcha challenge. By analyzing movement characteristics, we can distinguish between human users, who exhibit more varied and less predictable movement, and bots, which tend to have consistent and uniform patterns.

Approach

The approach for detecting bots versus humans is based on analyzing movement data captured during captcha interactions. The key steps in the methodology include:

1. Data Collection:

   • Movement data is collected, including X and Y coordinates that represent the user's interaction during the captcha challenge.

2. Preprocessing:

   • The raw movement data is cleaned and transformed into a format suitable for analysis, extracting key features that reflect user behavior.

3. Feature Engineering:

   • The following features are derived from the movement data:
     • Mean Position: Average X and Y coordinates.
     • Standard Deviation: Measures variability in movement.
     • Movement Speed: Captured by analyzing the differences in movement over time.
     • Variability Metrics: Such as counts of direction changes and constant movement detection.

4. Model Training:

   • A machine learning model (Random Forest Classifier) is trained using the engineered features and labeled data (human or bot).
   • The model is evaluated using accuracy metrics to determine its performance in distinguishing between human and bot interactions.

5. Results:

   • The model's predictions and actual labels are displayed, showing how well it can identify human versus bot behavior.

Getting Started

To run the application:

1. Clone this repository.
2. Install the required libraries:

   pip install -r requirements.txt
   

3. Run the Streamlit application:

   streamlit run mode.py