8.8KB
title: HackTheNorth 2017 – University of Waterloo author: rhnvrm type: post date: 2017-09-20T14:25:42+00:00 url: blog/2017/09/20/torogo/ categories:
- projects tags:
- datascience
- hackathon
- python
I attended HackTheNorth, which is Canada’s Biggest Hackathon and takes place at University of Waterloo campus. I was glad to join over 1000 students from around the world at the University of Waterloo to collaborate and create something extraordinary in 36 hours. I didn’t go in as a team as the visa process was a bit delayed and so none of the participants from India knew if they would be able to go. Well, eventually, we were three people from different universities from New Delhi who formed a team when we met in Canada.
We are excited to announce that Canadian Prime Minister Justin Trudeau will be giving the welcoming remarks at opening ceremonies tonight! pic.twitter.com/cWePNlFXYj
— Hack the North (@HackTheNorth) September 15, 2017
On the way to #HacktheNorth by the shuttle. Super excited! pic.twitter.com/T2L3mYzb03
— Rohan Verma (@rhnvrm) September 15, 2017
@JustinTrudeau kicking off @HackTheNorth ! Proud to be Canadian 🇨🇦 pic.twitter.com/3RaIHlJYWF
— Yingning (@yingninggui) September 16, 2017
Finally found where we will be hacking for the weekend at @HackTheNorth! Such an amazing space! #HacktheNorth pic.twitter.com/yxsWGv7oWr
— Rohan Verma (@rhnvrm) September 16, 2017
Let Data Science decide the perfect neighborhood for you in Torontohttps://t.co/LwisOhmlZV#HacktheNorth pic.twitter.com/0L9oIZVmfk
— Rohan Verma (@rhnvrm) September 17, 2017
The Project
We built ToroGo, an app powered by Data Science to help newcomers decide the perfect neighborhood for someone moving to Toronto. I worked on writing the API for the backend and the entire data collection and analysis involved in the project.
Inspiration
We traveled all the way from India and while booking our AirBnb we were overwhelmed with the options available to us. Although, websites like tripadvisor help with reviews of an area, it is often not enough. Toronto welcomes over 40 million visitors annually, and is the leading tourism destination in Canada. So we decided to use the power of data science to help travelers like us solve where to stay.
What it does
Find a Place
In our discussions we came up with three important things that users look at while deciding where to live in a new city:
- Price
- Safety
- Reviews
The app requests the user to enter his priority for the above mentioned features and suggests the top 5 suggested places that would suit the user based on data analysis.
Currency Converter
Along with that we provide the user easy access to check the live currency exchange rates using XE.com powerful api instantly from the app. It supports searching from 100s of currencies and can instantly give the live exchange rate.
How we built it
We used open datasets available from the Canadian Open Dataset website and Airbnb to build an aggregate score of around 130 sub-areas of Toronto City. We chose three parameters,
- Locality Ratings based on Sentiment Analysis of Reviews
- Safety Ratings based on Crime Data such as Assaults
- Priceyness Ratings based on the cost of listings of Airbnbs in a Neighborhood.
The data analysis can be viewed in our iPython Notebooks
After this, we wrote an algorithm based on Euclidean Distance between these ratings and the user’s preferences to suggest the most suitable locations for the user.
The currency converter is powered by XE.com‘s API and we were able to make an intuitive UX for searching through the currencies.
Challenges we ran into
- Data was easy to find but aggregation was tough to do based on the neighborhood level.
- Getting the heatmap to work based on neighborhood on Android
- We tried to integrate the Microsoft Azure Chatbot API but were unable to integrate it with the app.
Accomplishments that we’re proud of
- A full ranking of the neighborhoods based on reviews, safety and priceyness of Toronto
- A fully functional App and API.
- Firebase Authentication