How mukhilamsath/instagram-fake-account-detection---multi-class-classification works
FakeGuard AI classifies Instagram accounts as real, active fake, inactive fake, or spammer. The principal workflow is account-feature entry in the web UI, submission to FastAPI for feature preparation and model inference, then display of a class, confidence, and signals. The repository also includes a Streamlit exploration app and a separate dataset-to-model training pipeline. The README documents both interfaces; source excerpts show their main entry points, but not the template or CSS internals.
Main entry points: server.py FastAPI Server · main.js UI Interactions · app.py Streamlit Explorer · dataset.csv Account Dataset
Components 13
Web Scanner
- FakeGuard UI
templates/index.html - UI Interactions
static/js/main.js - FastAPI Server
server.py - Feature Preparation
server.py - Class Prediction
server.py - Dataset Stats API
server.py
Model Explorer
- Streamlit Explorer
app.py
Model Training
- Training Data Prep
train.py - Feature Engineering
train.py - Model Training & Evaluation
train.py - Model Artifacts
train.py
Other components
- Account Analyst
- Account Dataset
dataset.csv
Connections 17
- Account Analyst → FakeGuard UI: opens
README.md - FakeGuard UI → UI Interactions: loads UI
- UI Interactions → FastAPI Server: submits scan
static/js/main.js - FastAPI Server → Feature Preparation: parses features
server.py - Feature Preparation → Class Prediction: builds model input
server.py - Class Prediction → UI Interactions: returns result
static/js/main.js - UI Interactions → FastAPI Server: requests stats
static/js/main.js - FastAPI Server → Dataset Stats API: serves stats
server.py - Dataset Stats API → UI Interactions: returns metrics
static/js/main.js - FastAPI Server → Account Dataset: reads dataset
server.py - Account Analyst → Streamlit Explorer: explores models
README.md - Streamlit Explorer → Account Dataset: reads dataset
app.py - Streamlit Explorer → Model Artifacts: loads models
app.py - Account Dataset → Training Data Prep: loads samples
train.py - Training Data Prep → Feature Engineering: engineers features
train.py - Feature Engineering → Model Training & Evaluation: trains and evaluates
train.py - Model Training & Evaluation → Model Artifacts: saves models
train.py