Inspiration

As per the book "The Secret", maintaining positive aura / positive thoughts brings positivity to life. Today's media is selling too much negativity. This is an attempt to highlight the positive news first.

What it does

  1. It Aggregate news from leading Newspapers such as NYT, CNN, FOX etc..using their RSS feeds.
  2. Use classifier model (NLTP, TextBlob) and perform sentiment analysis to find the polarity of the article.
  3. Based on the output, positive items are displayed on top, followed by neutral and negative items.

How I built it

It's built using Python (Feedburner,NLTP,TextBlob,NaiveBayesClassifier model) along with MySQL, MongoDB, NodeJS and Express JS. Currently its hosted on AWS server.

Challenges I ran into

Train the model so it understands the difference between positivity and negativity. Experimented multiple cloud solutions (mLab, aws bringing them together was another challenge).

Accomplishments that I'm proud of

I'm proud of the entire project as I pulled the entire thing in 24 hours.

What I learned

Different ways to train the model (different libraries) along with Express JS framework.

What's next for Positive News First

Automate all the 3 modules and train the model on an ongoing basis.

Special thanks to Ajanni Haltiwanger for assisting me in collecting news items and testing the site.

Share this project:
×

Updates