The original project was focused on providing resources for people to figure out the safest roads to take while walking at night, using light density, density of public areas, and other factors to predict a safety rating through a machine learning algorithm for each road. However, we ran into issues with collecting sufficiently useful and precise data for this problem, so in a late night decision we took inspiration from older ideas that day and members' prior coding projects to attempt to make a Nativescript mapping application involving Twitter, Watson, and the election.
What it does
The app is intended to display a real-time graph of Twitter posts related to the election, with some indicator to convey the emotional parameters of each post, namely the extent to which each post demonstrates joy, sadness, anger, disgust, and fear. This data might be useful for anyone interested in possible correlations or trends between location and the emotional character of election posts.
How I built it
We used Nativescript to develop an iOS application for the project. We also employ the APIs for Twitter and Watson in order to collect and analyze data.
Challenges I ran into
As mentioned earlier, we ran into many problems with data collection for the original project, as we could not find all or process all of the data we needed in a reasonably precise form. We also had issues with getting Nativescript to install and run, but we were eventually able to figure out how to make this work.
Accomplishments that I'm proud of
We're reasonably proud of the ideas we were able to eventually come up with and our perseverance to continue in spite of the lack of extensive hackathon experience on our team and the odds against us finishing.
What I learned
It has most definitely been an experience for us to attempt to come up with and execute a project idea. We learned how to hackathon and how not to hackathon, especially that coming up with an achievable project is pretty hard.
What's next for Mapping Twitter with Emotion Analysis
Making it legitimately work, and then also making the original project work too.