Visit website


Every time we open our timelines, we're met with endless negativity and reminders of the situation we currently face.

What it does

Pulls in a selection of news articles from an API, filters out any blacklisted words and uses NLP to perform sentiment analysis on the title/description. Based on this, it determined how "happy" the article is, and filters out any articles deemed too negative.

How we built it

Using node.js and Glitch for rapid deployment, great teamworking environment & easy web-app hosting.

Challenges we ran into

With only the headline & first line of the article, it is difficult to get a full understanding of the nature of the article.

Accomplishments that we're proud of

The site is fast, responsive and uses API requests sparingly without hindering the user's experience.

What we learned

NLP is difficult to perfect!

What's next for Positive News

We'd love to be able to pull in from more APIs to get a broader range of articles. There was also discussion of a "Featured" sorting method where articles could be handpicked or even voted on.

Share this project: