Business

Hack the North Spotlights: Dropstats, crowd.data, Hive & VacAlert

Learn how four student teams used Dropbase for a variety of different solutions at Hack the North, solving problems across many industries.

In our final post of our Hack the North 2020 series, we wanted to showcase some of our favourite uses of the Dropbase API. We were fortunate to have a ton of entries to our API Challenge, so it was hard picking our favourites, but the projects below showed some very unique ways in which the Dropbase API could be combined with other technologies to create some very cool projects.

Web Scraping & Sports Stats

DropStats

For avid sports fans and sport professionals alike, tracking statistics about their teams is always a fun challenge. Finding ways to accurately track and creatively visualize statistics allows for greater insights into the underlying performance of the team, and can expose surprising correlations and statistical relationships.

Visualization of NBA scores using Dropbase and DropStats

About the team: The DropStats team was comprised of three high school students from Richmond, BC: Yifan Hao, William Kang, and Hana Domingo. This was their first time attending Hack the North, and only their second hackathon ever!

What the Tool Does: DropStats scrapes the NBA website after every Raptors game to get the individual player statistics, and automatically adds it to a database

How it's built: The team used Selenium to scrape the web, and then recorded the results using the Google Sheets API. Once the data was in the google sheet, the CSV file was sent to Dropbase to store the aggregated data. They then took that data and displayed it visually using Tkinter.

How they used Dropbase: Their main use was to store the web scraped data in a database, and then use that database to fetch the data for their frontend visualizations. When asked why they chose to use Dropbase, they said "because it seemed like an easy and efficient way to take the data we got from web scraping into a database"

Checkout the full project video:

Machine Learning and AI Dataset Crowdsourcing

In today's economy, most decisions are in some way made on the basis of data. From machine learning to artificial intelligence, businesses and researchers are creating new ways of analyzing data to optimize decisions and build powerful models. Getting the data necessary to create these models can sometimes be a difficult task: either the data isn't available, or it's so spread out that collecting data sets large enough to create valid models is infeasible.

There were two teams that attempted to solve this problem from different angles, and we felt that both of them delivered great solutions that we wanted to include.

Hive

Screenshot of Hive contribution screen

About the team: The team was comprised of Eshan Betrabet, a Computer Science student at Carleton University, Imaan Gill and Kanwarpal Brar, who both attend the University of Waterloo for Computer Science, and Yusuf Nissar, who is doing a dual degree in Business Administration and Computer Science at the University of Western Ontario.

What the tool does: The Hive platform allows users to build machine learning platforms in a collaborative manner, with the crowdsourcing of data collection. Users are able to make request for data they want, and other users are able to upload datasets that they might have that fit the criteria. From those CSV uploads, the data gets standardized and put in a database for further use.

How it's built: The project was built using Django and Firebase, with the Tailwind CSS library making up the styling for the frontend. To deal with the file uploading and processing and storage, this was done with a combination of python and API calls to Dropbase to store the data.

How they used Dropbase: Dropbase was used as a way to combine and clean datasets that were provided by multiple sources. When users uploaded data to contribute to a project, the data was then sent to Dropbase, where preprocessing steps were performed, and the data was uploaded to a database that the user who requested the data could then access.

Checkout the full project video:

crowd.data

crowdsourcing data requests and data uploads with crowd.data and Dropbase

About the team: The crowd.data team was made up of three Computer Science students from the University of Waterloo, Michael Le, Franco Chen and Kevin Lu, alongside their friend George Saad, who is currently completing his undergrad in Engineering Science at the University of Toronto

What the tool does: They created a platform that allows users to contribute smaller datasets that could then be aggregated into a larger, unified dataset that people could use to train models on. To do this, they created a way that users could request data, and a way for users to contribute data, and a way for users to comment and add feedback about data that was requested or provided.

How it's built: The team built the application using React for the frontend of the site, and Node for the backend. They used CockroachDB as their database, which used to store everything from their users to the contribution data. Individual users were able to either post a request for data, in which case they also entered in a Dropbase API Key as to where the data would be aggregated and cleaned, or they could upload their own datasets to threads already created, which would then be sent to Dropbase to be cleaned.

How they used Dropbase: For the team, it was important that they had a way that users could combine datasets in a manner that made the datasets consistent, and were able to be uploaded to the same source. Dropbase allowed users to upload CSV's and automatically convert them into a database format that they then cleaned, and then sent to their database to be aggregated with the other contributions that were made on the request thread.

Checkout the full project video:

Vaccine Communication Systems

VacAlert

Determining patient eligibility and locating vaccination sites with VacAlert and Dropbase

About the team: VacAlert was created by brothers Justin and Nathan Chung, who are currently in high school at The Woodlands Secondary School in Mississauga, Ontario. It was Justin's second time at Hack the North, and Nathan's first.

What the tool does: VacAlert is a personalized vaccine tracker that provides the latest updates and simplifies the process for vaccination. It notifies people of their changing eligibility and timeframes for receiving the COVID 19 vaccine. It checks official government sources for any recent news, and alerts users in real-time if any news has an impact on when they will be receiving the vaccine. For example, if a vaccine shortage is announced, or if a new profession is covered under the  first wave of vaccinations.

How it's built: They built the frontend of the application using Angular and material design principles. For the backend, they built with IBM's LoopBack 4 framework. For their database, they used a CockroachDB populated with data that they web scraped data and then cleaned via the Dropbase API. They also integrated the application with the Vonage API to send SMS notifications to suers, and push notifications to users through the Firebase Cloud Messaging API.

How they used Dropbase: The team needed an efficient method of processing and updating their database on a regular basis, with the data that they scraped from government sources. They used the data pipelines to automate this process, and then used the postgREST API to interact with the processed data for their application.

Checkout the full project video:

Interested in finding out how Dropbase can supercharge your project?
Sign up for free today, or contact us to schedule a demo

Newsletter
Insights and updates from the Dropbase team.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
By signing up you agree to our Terms of Service