NewsMap

Inspiration

With the media's exponential rise in shaping our opinions, we are often constrained in our understanding by our limited media intake. We wanted to remedy this problem by providing an easy-to-use web app, that would allow people to gain a global perspective on the media attitude toward critical issues. We began by focusing on companies, since attitudes toward companies vary significantly across the world, and can often lead to friction.

What it does

NewsMap utilizes global market visualization software to provide a real-time visualization of news reports mentioning companies across the world in over 65 languages. By customizing the tone and geographic data from these news reports, NewsMap provides simple, easily understood, and interactive graphics, representing historical and up-to-date data, along with a company's stock and financial performance metrics.

NewsMap allows companies to quickly get fast, and reliable data on their international digital footprint, empowering them to make strategic and rewarding decisions.

How we built it

We used a combination of Python, HTML, Javascript, CSS, along with the MapBox API and Carto API to create our interactive data visualization web app. We obtained our datasets from GDELT, an online database containing metadata on articles published in over 65 languages. We wrote scripts to download, parse, and analyze the metadata to provide meaningful results based on a user's search query. These results were fed into Carto and MapBox (through a Flask - a Python micro framework), to visualize the data, and provide a web interface for a user to interact with.

Challenges we ran into

It took us a while to realize that the url handled in flask needs to be in a static folder unless the file cannot be correctly accessed. Formatting and parsing the structure of geoJSON files proved to be a difficult and meticulous task. This is also the first time any of us has ever coded in Javascript and it took many failures and setbacks, combined with multiple calls for help from mentors in order to get a fully functional JavaScript code. All in all, there were many hurdles and a complete all nighter to get to where our web app is today; however, it was well worth it because of how much we learned.

Accomplishments that we're proud of

We are proud to have a fully functioning web app with our desired fundamental capabilities that we set out to create. It was a phenomenal experience having the opportunity to work collaboratively in a team, having both trust and confidence in one another to help each other, and build a cool product.

What we learned

We learned a lot as we are all pretty new to hacking; one of us was in fact a first time hacker! We understood how to code in HTML and Javascript to create an interactive web app and also work with APIs and large datasets to visualize and interact with data. At the same time, we also learned to embed python scripts into HTML code using flask. At the same time, we also learned many tips in debugging and learning about all the steps in growing from an idea into a fully functional prototype.

What's next for NewsMap

We plan to incorporate a randomly weighted majority algorithm (Expert's Algorithm) in our code. It is a simple but effective machine learning algorithm that can allow for supervised learning for our program, "teaching" it the opinions of which areas in the world have higher influence on the stock prices or stock volatility of companies, helping learn which areas they should allocate their resources . We also hope to expand our web app's computational power, expanding its visualization to longer time spans. We also hope to reduce the run time of our code and increase our capacity to handle larger data sets.

Built With

• carto
• mapbox