/Sentiment-Analysis

Sentiment Analysis of Youtube Comments

Primary LanguageJupyter Notebook

Sentiment Analysis

Sentiment Analysis of Youtube Comments

presentation | pdf

To Check

chromedriver under webdriver directory is currently for chrome version 86 Mac. If you are using different version of chrome or different OS, you should download suitable chromedriver from this link.

How to run this web server

  1. Download data input
    Download data of twitter dataset with sentiment from this link
    Place this data in an input folder in root directory

  2. Build model
    Open model.ipynb and execute all cells
    model.h5 and tokenizer.pkl will be created

  3. Run Web Server

python3 app.py
  1. Open Web Server
    Open webserver by entering address below on browser
127.0.0.1:5000
  1. Analyse
    Enter youtube url on the web server page. The program will automatically collect comments from that youtube video and perform sentiment analysis on them. The results will be shown as pie chart.

Reference

Flask referenced from this github
The method to build model was referenced from this link