shreychauhan02/comment_sentiment_analysis_project

How shreychauhan02/comment_sentiment_analysis_project works

This is a YouTube-comment sentiment product combining a Chrome extension, a local FastAPI analysis service, and an MLOps training pipeline. The user opens the extension on a video; the backend retrieves comments, analyzes sentiment and spam, and returns insights for display and CSV export. The documented training stages progress from ingestion through registration, with MLflow/Dagshub tracking. Source excerpts support the API’s YouTube and Gemini calls and the pipeline modules; some stage links and the extension-to-backend relationship are documented in the README rather than shown in sampled calls.

Main entry points: main.py Analysis API · utils.py Pipeline utilities · model_building.py Sentiment training · model_evaluation.py Model evaluation

Components 14

User Experience

Extension UI extension/popup.js

Analysis Service

Analysis API fastapi-backend/main.py
Spam training fastapi-backend/train_spam_model.py

Model Pipeline

Data ingestion src/data/data_ingestion.py
Text preprocessing src/data/data_preprocessing.py
TF-IDF features src/data/feature_extraction.py
Sentiment training src/model/model_building.py
Model evaluation src/model/model_evaluation.py
Pipeline utilities src/data/utils.py

Model Operations

Model registration src/data/register_model.py

Other components

YouTube viewer
YouTube comments
Gemini summary
MLflow / Dagshub

Connections 21

YouTube viewer → Extension UI: opens README.md
Extension UI → Analysis API: requests analysis README.md
Analysis API → YouTube comments: fetches comments fastapi-backend/main.py
Analysis API → Gemini summary: requests summary fastapi-backend/main.py
Analysis API → Extension UI: returns insights README.md
Spam training → Analysis API: trains spam model
Data ingestion → Text preprocessing: feeds pipeline README.md
Text preprocessing → TF-IDF features: feeds pipeline README.md
TF-IDF features → Sentiment training: feeds pipeline README.md
Sentiment training → Model evaluation: feeds pipeline README.md
Model evaluation → Model registration: feeds pipeline README.md
Data ingestion → Pipeline utilities: uses src/data/data_ingestion.py
Text preprocessing → Pipeline utilities: uses src/data/data_preprocessing.py
TF-IDF features → Pipeline utilities: uses src/data/feature_extraction.py
Sentiment training → Pipeline utilities: uses src/model/model_building.py
Model evaluation → Pipeline utilities: uses src/model/model_evaluation.py
Model registration → Pipeline utilities: uses src/data/register_model.py
Sentiment training → MLflow / Dagshub: logs run src/model/model_building.py
Model evaluation → MLflow / Dagshub: logs metrics src/model/model_evaluation.py
Model registration → MLflow / Dagshub: registers models src/data/register_model.py
Analysis API → MLflow / Dagshub: loads registered model fastapi-backend/main.py

Drawn by GitDiagram on November 12, 2026 from the repository's file tree, README and source files. Read as Markdown ·