Thrifty for Amazonians

We first started out wanting to automate the downloading process for students who use Blackboard. We considered the two ways to accomplish this task. First, we considered using the Blackboard API to access the account of a student and automate their downloading process. However, we quickly found that it was really difficult given our time frame to learn how to use it with its limited documentation. Second, we considered doing web scraping to access the data. It was a good idea until we realized that this project wasn't something that could be done within our given time frame as well. As such, we branched off to a new idea. We found that amazon was an easier website to scrape through, so we came up with the idea of creating an app which can be used to find thrifty prices for similar products.

What it does

The app notifies user's if there is a good deal on an item they are looking to buy. So, you just type in your item and it will search amazon and tell you if there is a similar match.

How I built it

When the user starts the app, the user will be prompted to search for any item on Amazon. By web scraping through Amazon, we parse through the html and look for specific patterns that will return the title, price, img_url, and URL. With that data, we use the item's name to create categories. We the price of every element with that matching category into a list. Then we use that to figure out the standard deviation on the category and if there is an outlier less than the mean we notify the user

Challenges I ran into

When web-scraping the data from Amazon we found out how different the HTML of each code can be so it becomes difficult to find regular expressions to parse the data. I also learned how hard is to try and create your own machine-learning algorithm. Also, I learned to allow for more time when planning projects.

Accomplishments that I'm proud of

I am proud that we have built something that is semi-useful and interesting in 24 hours. We learned about automating things using web-scraping which is an awesome skill. I learned to work with other people to balance out my weaknesses and got reminded about the satisfaction and the patience it takes to work successfully in a team.

What I learned

I learned about automating task through web scrapping.

What's next for Thrifty for Amazonians

In the future, we will be implementing a feature that will track the item on behalf of the user and email them if it finds a deal. Furthermore, we will utilize the information that the bot has been gathering in its current phase to make it more efficient. So, it is already storing the categories and their respective variance as well as the size of the sample. This will help it make a more educated guess about a good deal on categories. Also, our next step is to build a web application for our app so more users can access it. Another thing we are looking to do is to optimize our algorithm by refactoring our code and using multiprocessing.

Built With

• beautiful-soup
• bs4
• mongodb
• pymongo
• pysimplegui
• python
• selenium