NOTE: This repository is for archive purposes and will not be updated. However, you may still find some useful code here for your own projects
Stock sentiment analyser. Scrapes news articles using NewsAPI, Twitter sentiment for a stock using SocialSentiment.io and sub-reddits posts (and comments if wanted) from Reddit. News articles, tweets and Reddit posts are then analysed for sentiment using VADER Sentiment Analysis. A composite sentiment score is calculated and plotted with stock price for the last month.
This was created a small personal project to learn how to scrape different sources and use sentiment analysis. Comparing social media sentiment against stock price was an interesting project to learn these skills.
- Get a free API key from NewsAPI
- Get a free API key from SocialSentiment.io
- Follow this tutorial to set up an app in Reddit to get an app ID, app secret, app name. Use these details with your reddit username and password. This is needed to be able to scrape content from Reddit.
Copy config.yaml.example to config.yaml and enter your API keys for SocialSentiment, NewsAPI and the details for Reddit. Also define the sub-reddits you want to scrape.
Create a virual environment python3 -m venv venv and activate the environment source venv/bin/activate. Install the requirements pip install -r requirements.txt.
Run the dashboard python3 stock_sentiment_analysis.py and go to localhost:8050 in your browser. Then enter the ticker and wait for the scraping and sentiment analysis to complete. Note that it can take a few minutes for the analysis to run.