Machine learning pipeline for end-to-end Analysis of Tweets of user using NLP models

Required python version: 3.7 or higher and MongoDB Installed Locally.

pip install -r requirements.txt

1. Build docker:

docker build --tag tweet_analysis:latest .

2. Run docker server:

docker run -d -p 8080:8080 tweet_analysis:latest

3. Use Docker Compose: It will create both the images (mongoDB and fastAPI) and start the containers.

docker-compose up --build

uvicorn run_server:app --host "0.0.0.0" --port 8000

For Visualizing application Visit: http://localhost:8000/docs

• /health Endpoint for checking server availability.
• /user/signup Endpoint for registering a user.
• /user/login Endpoint for login a user and providing an access token.
• /view_tweet/ Endpoint for getting a tweet from its Id.
• /view_all_tweet_ids/ Endpoint for getting all ids of tweets stored in DB along with Its User Name.
• /extract_tweets/{user_name}} Endpoint for extracting loading and transforming tweets from Internet.
• /analyze_sentiment Endpoint gives sentiment against a tweet by providing a Tweet ID.
• /analyze_emotion/ Endpoint gives emotion analysis against a tweet by providing a Tweet ID

pytest .

Project file structure is:

.
├── tweet_analyzer
│    ├── __init__.py
│    ├── pipeline.py
│    ├── models
│    │    ├── __init__.py
│    │    ├── base.py
│    │    ├── sentiment_classifier.py
│    │    └── emotion_analysis.py
│    │  
│    ├── utils
│    │    ├── __init__.py
│    │    ├── db_utils.py
│    │    ├── fastapi_models.py
│    │    └── parse_config.py
├── test
│   ├──__init__.py
│   ├──test_parse_config.py
│
└── docker-compose.yml
└── Dockerfile
└── config.yaml
└── README.md
└── etl.py
└── server.py
└── requirements.txt
└── model.py
└── .env

For adding the authentication to fastapi endpoints, code was taken from following URL

https://github.com/BekBrace/FASTAPI-and-JWT-Authentication

Sibtain Raza