Introduction to the News-Map.io
Overview of articles' sentiment in northeastern USA
Overview of articles' sentiment in midwestern USA
Misinformation is a prevalent problem in the news industry. Most people are isolated in their own bubbles of thought. Our solution to address this lack of perspective is to better inform readers of current events across the world from various points of view. Providing readers with lenses on events allows them to see the full story as well as the biases that they otherwise would miss. Informing the people of a source's interpretations would allow people think for themselves.
What it does
Aggregates different news sources and analyzes recent articles to provide users with the ability to recognize location-based trends in current events and summarizing each sources view as either positive or negative.
How we built it
Backend and APIs
Python, Flask, NYTimes API, IBM Watson Natural Language Understanding API, News API, BeautifulSoup
Challenges we ran into
Parsing and cleaning data from various news sources, some even having multiple page organizations, forced us to manually scrape each news source. Retrieving location in articles' body to visualize was challenging since we couldn't possibly account for every possible location and had to think intuitively to extract these datapoints.
Accomplishments that we're proud of
Our frontend provides users with informative and engaging visualizations of the current events, allowing them to see much more than just a few sensational headlines or polarized perspectives.
What we learned
How to extract key information, such as location data, from articles' contents as well as extracting article bodies from webpages by scraping.
What's next for News-Map.io
Improving the sentiment analysis of the articles. Removing repeated code blocks to follow the DRY principle and improve efficiency. Engaging with more sources for more diverse perspectives.