Inspiration
At first we hoped to use a tool for algorithmic trading; however, we realized it would be more impactful to create a tool that can display the global news sentiment for any search query (not limited to trading).
What it does
Displays the sentiment of news articles for any search query. Scrapes news.google.com results for top articles related to a query and sorts by sentiment. Global trends will be able to track the sentiment of the top stories section for the news.google.com site in each of the listed english speaking countries (necessary for sentiment analysis).
How we built it
Flask, Python, Web Scraping (Beautiful Soup, requests), lots of blood, sweat and tears :(
Challenges we ran into
Accessing current articles and scraping data (no APIs were working, had to scrape different formats directly from news.google.com, designed our own scraper), getting the website to format correctly, integrating everything into one cohesive web-app
Accomplishments that we're proud of
Scraped lots of info in real-time and can provide sentiment (polarity/subjectivity) for every single article. Overall, we're happy that we can provide a resource to help people around the world better understand the information that they consume.
What we learned
Everything
What's next for Senti
- Global Trends Page
- Trends Over Time for User-Chosen Queries
Log in or sign up for Devpost to join the conversation.