Sentiment Analysis of YouTube videos with Gcloud NLP API

Automatically download and analyze any YouTube video subtitles based on sentiment with Google's Natural Language API. Exports a CSV with score, magnitude, and overall sentiment and a transcript .txt file. Sentence query system is a simple fork of Gclouds NLP API tutorial for Sentiment Analysis.

Made for the 'PodPlot' assignment for UC Berkeley's 'Technology Design Foundations'.

Follow this tutorial for authentification: https://cloud.google.com/natural-language/docs/sentiment-tutorial?authuser=2

Make sure that you have satisfied the prerequisites mentioned in the tutorial above and that the GOOGLE_APPLICATION_CREDENTIALS variable shows the path to your JSON credentials.

Usage

python3 sentimentAnalysis.py "YouTube URL"

Example with the 'The next outbreak? We’re not ready | Bill Gates' TED talk

python3 sentimentAnalysis.py "https://www.youtube.com/watch?v=6Af6b_wyiwI"

Installation

Clone this repository, cd into it, and install the required packages with pip.

git clone https://github.com/titusss/Sentiment-Analysis-with-Gcloud.git
cd Sentiment-Analysis-with-Gcloud
pip install -r requirements.txt

Case Study

Example infographic created with Google Spreadsheet Charts.