This python module gets tweets based on keywords from twitter and does some basic analyzes on them. This microservice was part of a bigger project and used as a helper to the main module.
This module follows three steps: 1- Getting a certain amount of tweets based on pre-defined keywords. 2- Getting one page of the latest tweets of users. (timeline) 3- Run analyzes 4- Save in the database
1- Twitter API (tweepy) 3- AsyncIO 4- RabbitMQ 5- MongoDB
You can change the configs via docker-compose.yml or if you want to run them directly just take a look at config.py.
# Twitter API access configuration
TWITTER_API_KEY:
TWITTER_API_SECRET:
TWITTER_ACCESS_TOKEN:
TWITTER_TOKEN_SECRET:
# RabbitMQ Configuration
# You can leave it alone but be noticed you need to set the same value to the rabbit service as well.
RABBIT_USERNAME: admin
RABBIT_PASSWORD: 123123
RABBIT_URL: rabbit
# MongoDB Configuration
# You can leave it alone but be noticed you need to set the same value to mongodb service as well.
MONGODB_USERNAME: dbuser
MONGODB_PASSWORD: dbpassword
MONGODB_URL: mongodb:27017
MONGODB_DBNAME: tw_analyzer
# These are the comma-separated keywords that the module will look for on Twitter at starting. It's for the best to not exceed 10 items.
SEED_KEYWORDS: apple,microsoftThe concept of this helper microservice is to run two analysis:
1- Polarity Polarity will show the positive and negative rate of a text. There return if a numeric float between -1 to 1. The lower shows more negative and the higher show positive. 0.0 also shows natural.
2- Subjectivity It's a float between 0.0 to 1.0 that shows the subjectivity or objectivity of a content. The 0.0 shows the content is very objective and 1.0 is very subjective.
docker-compose up --build docker-compose down