My friend is a trader who got me interested in creating trading algorithms. I wanted to be better at looking at how the world operates financially, economically, culturally, etc. There's way too much data for any person, so our team sought a way to compress all of that available information.

What it does

It's an analytics tools that shows how a country or city is related relative to other countries or cities. For example, it can show you the relationships between Canada's and other countries' economies, cultures, etc. But at the same time, the user can choose to change to a different point of reference, like China or Japan.

For now, all it does is examine news sentiments between a few related countries, but in the future it will be able to handle a plethora of data, like tweets and blockchain data, from all countries and major cities. The data will be used to help nations with their development strategies, better educate people, and more.

Psychological studies shows that the media nowadays often acts as an echo chamber and people tend to have a 'confirmation bias' that reinforces previously held beliefs and filters out information that contradicts their beliefs. We hope that our project can help individuals overcome this issue and develop a 21st century common sense because understanding conflicts, social problems, and the status of the world should be the responsibility of an educated citizen of the world.

With our project, we hope that people can understand geopolitical tensions, development of transnational corporations, and various perhaps sensitive cultural issues to help them make informed political decisions such as voting, economical decisions such as investment, and decisions in their day to day life such as choosing to thoroughly discuss or completely avoid sensitive political issues. Perhaps this project can give students like us inspirations for ideas for the next hackathon!

How I built it

I created a Python program to automate news scraping. I sent the news to Azure's text analytics tool to get sentiment. Sentiments and news data are sent to Firebase for access to our client app (a dashboard). The news collector is made with a Python script that runs through CRON, and the client is made with HTML, CSS, and vanilla JavaScript.

Challenges I ran into

  • Frontend skills are not that great, so that's something to improve upon.
  • We ran out of credits with the news API and Azure
  • Defining the project scope

Accomplishments that I'm proud of

  • Commutative hashing for data storage
  • Usage of a geographic heatmap

What I learned

  • Google cloud SDK set up for Python
  • Team members learned more about Python and JavaScript

What's next for Mountain

  • A rewrite with a frontend Framework
  • More product design and research
  • Creating our own machine learning models
  • Developing algorithms to reduce bias in news
  • Enabling searches to more countries and better smart filters for robustnest
  • moving beyond news -- stock market updates, twitter, general background, immigration patterns, etc
Share this project: