Inspiration
Nowadays, when searching the housing market, we often have to view multiple listings. This often gets tedious and time consuming very quickly. While researching ideas, we came across many different captchas. Filling out captchas was boring. As we scrolled through multiple websites, we filled out many different captchas. We thought about the idea of combining captchas and houses for sale using artificial intelligent.
What it does
Through the use of machine learning, it figures out houses that are similar to the house images the user selected based on patterns. It then recommends a list of house listings similar to user's preferences.
How we built it
We used React for our frontend. Flask was used for our backend. We used Javascript to web scrape.
Challenges we ran into
One challenge we ran into was web scrape listings from Zillow. While attempting to web scrape, we quickly realized that Zillow has security measures to prevent us from web scraping.
Another challenge we faced was some of us had no prior experience of full-stack web development.
Accomplishments that we're proud of
After we failed to web scrape Zillow, we quickly were able to pivot to Javascript. This pivot allowed us to access address and image urls.
We were able to convert images into byte arrays which we were able to send to our React application using Flask API.
We were able to create fully function web application using React.
We used deep image search library to find similar houses given user selected houses.
What we learned
We learned how to use web scraping, Rest API, React, Flask, and most importantly teamwork. We learned how to implement Python scripts into web application.
What's next for Captcha Yo House
Some features we plan to implement are selecting attributes such as by zip code. Another feature would be to continuously update house listings in the database. We wanted to implement a Tinder styled swipe system to find houses you like.
Log in or sign up for Devpost to join the conversation.