Our inspiration came from Starting From the Bottom of our commute here from GTA and Waterloo, as we ran thru the 6ix we realized that we had a lot of woes and that we did not Connect with ourselves well. So we decided our Motto for this Summer Sixteen Hackathon would be to create a webapp to Know Yourself better.
What it does
Upon Successful Facebook Login, our Webapp generates a “virtual you” using your facebook footprint and utilizes indico.io’s advance text and image APIs to perform sentiment analysis.
First, we apply sentiment analysis on your posts to determine your least and most positive moments, as well as your general happiness. We also classify posts and images via indico’s emotion analysis and map them to topics such as Fear, Anger, Happiness and Sadness.
Using indico’s text tags and keywords API, we generated a list of interests based on your posts and pages liked. We also calculated your popularity over time by measuring the number of likes on your profile pictures. Using google’s chart API, we were able to graph and visualize this data.
Using Indico’s people processing API and facebook’s graph API, the app also calculates your top 5 closest friends.
On the image processing side, we used indico.io and Clarif.ai to proccess the facebook photos that you were tagged in. From your photos, our app is able to suggest the top 20 common keywords that resembles the overall theme of your photos, as well as categorize your photos based on the emotions and general atmosphere displayed.
How we built it
Of course, The Language that we used had to be the mighty, the best - node.js. The Views and the Controlla was built with Angular, and the backend was built using Express and Node.js. Indico API was Too Good in helping us Take Care of the Artificial Intelligence and Machine Learning of our Hack.
Challenges we ran into
There was Back to Back issues that we had to Catch up from. We tried to pull your full friend list from the Facebook API but we were Doing It Wrong as Facebook depreciated that feature.
Another challenge that persisted was trying to optimize the number of API calls we had to make in order to stay under our call limits - especially for Clarifai.
We are Proud of our accomplishments in using the Indico and Facebook API as we were able to get some of the Best Analysis of my Profile that I Ever Had. We were also proud of our Fancy Charts and our most popular topic ranker was pretty Hype.
What we learned
We learned that the Facebook API is Charged Up with features that can help create a webapp that can make a detailed summary of my posts All for Me
What's next for Know Yourself
We finished most of our hack at 5AM in Toronto and we're Faithful that once it was "Over", we could continue to add features such as your Worst Behaviour, and things you are Used To doing everyday. We also wish to further apply sentiment analysis by displaying long term trends of your mood. We hope that this webapp will be Unforgettable and receive prizes and Trophies.