Inspiration

We wanted to illustrate social relationships in a creative way, while learning to use new API's such as IBM Watson and GMail.

What it does

Reads your GMail messages with a particular contact, sends those messages for sentiment analysis to IBM Watson, and renders a visualization that mimics contacts (planets) orbiting around you (the sun). The distance from you to each planet is proportional to your sentiment score with that contact.

How I built it

By first logging in a user and asking them to add a contact. Then we make an API call to GMail to read email messages, parse those messages into a single string, and send that to another API call made to Watson. We then attach the score from Watson to that contact in our data model. We manipulate that score (which comes back as a number between -1 and 1) and use that to determine a planet's distance from the sun (the user).

Challenges I ran into

Gmail API is satan. Learning D3 was tricky.

Accomplishments that I'm proud of

Conquering the slime belching abomination that is google APIs. Also our d3 visuals are 1337.

What I learned

How to work with a team using the same GitHub repo, how to parse emails of different mime types, how to tie multiple different complex APIs together to achieve a single purpose.

What's next for SaturnV

Adding more features for visualizing user relationships, i.e. allowing users to click on planets to retrieve a graph that shows fluctuation in email sentiments with that contact over time.

Share this project:

Updates