Posted by on Feb 1, 2015 in | 0 comments

QuickChoice by Shashank AgarwalWe all hate making choices, where to go , what to eat , what to listen etc.

  • Used Facebook data of users to train machine learning & data mining algorithm, and then use these to recommend places like restaurants, music or movies.
  • Won First prize at “HackGT”, Georgia Tech for “Best Hack that Affects the Physical World” and was selected as one of “Top 10 hacks” out of 700 students.

Utilized Yelp API, Data Mining( Python-scikit), AlchemyAPI, Facebook API, Google Places API, Twilio API and php.

Wouldn’t it be great if an app just knew you too well and can suggest you what exactly you need. So here comes : “Snap Choice” . An app that learns about you from your facebook profile and uses advanced machine learning & predictive analysis to better understand what you like, what you don’t like and what you usually do.

We mine your facebook data and convert than into concepts or keywords. These keywords are added to your personal Data Mining Training Set. Now as you start using the app we learn from the way you use the app. We add all the paths you took in making a decision such as selecting a restaurant, listening to a song or looking for a place to go to.

The app has 3 main parts: 1) Travel 2) Listen 3) À la carte (Eat)

We use predictive analysis tools to find relations and predict the place you would love the most.

For Example:

You want to eat some spicy food. The app’s predictive analysis engine will fnd pattern and during our tests returned: {“food and drink”:”0.498103″,”cuisines”:”0.498103″,”mexican cuisine”:”0.498103″,”chinese cuisine”:”0.297418″}

So it categorized the keywords and found related items and we all know “mexican” & “chinese” is spicy food.

The same way we predict your travel queries. For Example you want to go to a “travel+food + beer” place. The app will understand what you mean and will give you a bunch of places you could go to.

Songs are the most coolest thing, wouldn’t it be awesome if the app always knew what you like and want to listen. So the app does that only and and uses facebook data to understand what you like and predicts the music you like.

Example: If you likes pages of ‘Justin Timberlake’ , ‘britney spears’ & ‘Justin Beiber’, the predictive analysis engine gives: {“art and entertainment”:”0.872367″,”shows and events”:”0.872367″,”concert”:”0.872367″,”celebrity fan and gossip”:”0.350339″,”music”:”0.158056″,”singing”:”0.158056″}

Which means it understands what you are actually looking for.

Does this sound cool to you? Lets take it up a notch, we have created API for all our services, especially the predictive analysis engine.