Skip to content

Latest commit

 

History

History
43 lines (35 loc) · 1.4 KB

README.md

File metadata and controls

43 lines (35 loc) · 1.4 KB
Raw

Sentiment Analyser Bot

What is it?

A chatbot, which analyses polarity of recent tweets of some given twitter handle. It classifies sentiment of the tweets into negative, neutral or positive calibre. (replies in the range of -1 to +1)

How it works?

  • The core of this project lies in the trained ML model of TextBlob module, which does our amazing work of Natural Language Processing of tweets.
  • Tweepy fetches required tweets at real time.
  • Flask handles the backend of our chatbot.
  • Ngrok hosts our bot on https server temporarily.

Prerequistes:

  • Python 3.7
  • Flask
  • Twitter API
  • Tweepy
  • TextBlob
  • Telegram bot API
  • Numpy
  • ngrok

How to create your own bot and use it?

  • Clone this repository
  • Download ngrok into this project directory: https://ngrok.com/download
  • Add your twitter API credentials in twitter_credentials.py file
  • Now, fire up the terminal and type the following:
username@username-PC: ~/myrepository$ ./ngrok http 5000
  • Add your telegram bot token in config.py file
  • Add your ngrok https URL (as obtained in the terminal output) in config.py file
  • Run the Flask server:
username@username-PC: ~/myrepository$ python3 app.py

There you go !!!

Go ahead, play with your bot!