/tweet-and-product-marketing

Analyse and Visualise the tweet sentiment on given product

Primary LanguageCSS

Product's Twitter Sentiment

Realization of the Smart tweet brief

Installation

  • pip install requirements.txt

Project structure

/configuration

  • resources.py: contains the key authentication of Azure API and Tweet API

/database

  • db_access.py : python code files for database manipulation. Provide connection to the database with some methods:

    • add_one_tweet
    • add_many_tweets
    • get_tweet_by_id
    • get_tweets
    • get_tweets_paginated
    • update_tweet_by_id
    • update_tweets
    • delete_tweets

    method call:

    from database.db_access import DatabaseManager as db

db.getInstance().add_one_tweet({
    'name' : 'tweet1',
    'sentiment' : 2,
    'text' : 'random comment'
})

/dataviz

  • SQL code
  • Python code for dashboard visulization

/datatweet: contains python classes that collect tweets with Twitter API and predict their sentiments with Azure API

  • tweet_manager.py
    • TweetCollection class: retreive most recent tweets
    • TweetSentimentPrediction class: send tweets to Azure in order to obtain their sentiment score and the confidence scores
    • TweetLoader class: prepare the the availability of database by calling TweetCollection and TweetSentimentPrediction and create the time series chart
  • tweet.py: an python object who transforms python object into json

Launch

  • python main.py

Credit

This project can not be completed without the help of Andrey and Mathieu