Newsbeat

Vision

Newsbeat shall become the number one source for realtime trend detection in leading news rooms across the globe.

What it does

Newsbeat visualizes the heartbeat of news. The web application takes the most recent news from Reuters and calculates the online resonance in realtime. The results are visualized in a compelling form on a worldmap.

How we built it

We fetch raw news data from Reuters global newsfeed, compute its resonance in social networks and plot it to an interactive worldmap. On a technical side we use Java to process the data, Twittter for calculating the online resonance and Google Maps for a stunning visualization.

Challenges we ran into

Limited API quotas forced us to implement workarounds and counter measures which we did not foresee. Insufficient quality and completeness of the source data lead us to reevaluate our options within a short timespan.

Accomplishments that we are proud of

We have built a working prototype, with which we validated our assumption that the resonance of news changes over a period of time. An earlier prototype allowed us to detect that relevant data was insufficient to accomplish one of the initial aims and we therefore had to pivot.

What we learned

On multiple occasions we experienced that the theoretical entropy of expected data is not necessarily matched by real data. As an example: Just because tweets may contain location information does not mean, that the majority of tweets does.

What's next for newsbeat

We would like to add more sources to calculate an even more accurate resonance indicator. To make the tool more useful for our potential customers, we would like them to customize the software to their needs. An additional feature we want to realize is the tracking of a single news story's relevance over time.