Thumbs Up News

What & Why

Thumbs Up News aims to filter news from different news publishers and show only positive or neutral news.

This project was created when the COVID-19 pandemic started and all you could see daily was bad news, it was pretty hard to filter news to see only positive ones and around this pandemic period, it seemed even more important to try and filter the news articles to try and get only positive ones.


We are using the RSS Feeds from news publishers to get our headlines, then we run the classifier to get a classification of the headline to see if that headline is a positive or a negative one. Then we put the headlines into the database and use Django as the backend and Nextjs as the frontend.


Our sentiment analysis classifier is not perfect and is currently running on a 81% accuracy, which means that there could be some false positives. To train the sentiment analysis classifier we used the python NLTK library.

To train the sentiment analysis classifier we used the following resources:

Our categories classifier is currently running on a 59% accuracy. The classifier was trained with different amounts of data on each category so the data fed to the classifier isn't balanced and that's why the accuracy is affected. We will be feeding more data to the dataset used and then retrain the classifier once we have a large enough dataset.

To train the category classifier we used the following resources:


For the scrapper, we are using the python library Scrapy to read the RSS feed and save the data from the feed. We are collecting the title, description, category and source.

We are using the RSS feeds from the following news publishers:


Thumbs Up Projects uses Django for the backend and uses the Django Rest Frameworkto generate the API to be called by the frontend. The frontend uses NextJS.

We are using Docker to keep the whole project in a container and serve it from a Digital Ocean droplet.


This project was created by FabioRosado a flight attendant that's learning how to code and hoping to do the jump into a developer position in the future.