Digital Advisor

Inspiration

After 3 semesters of course registration we wanted to find an easier way to pick out which courses we would need in order to fill out our graduation requirement. So we decided to improve upon the BU Course Search Tool, so instead of just giving the user a list of classes with the required credits, the Digital Advisor will plan out a whole schedule tailored to the user's major and desired hub credits.

What it does

Our website takes in the User's information such as their phone number, taken classes, and desired classes and uses a searching algorithm to create a list of courses that will fit the user's desires. The data that we use the search algorithm on is scraped from the Boston University Course Search Tool website and transferred to firebase in order to improve the time efficiency. From there the website makes an HTTP Request to the firebase data in order to determine what courses what fit the user's requirements. Finally after generating a list of the advised courses, using Twilio it sends that list directly to the user's phone via text.

How we built it

Our website uses Beautiful Soup to scrape data from the BU Course Search Website and sends that data to Firebase to be reused whenever we need it to. The website front-end created from HTML, CSS, and JS gets the user information and converts it into a JSON FILE that is sent to our API created from Python and Flask. The backend of the website then reads the JSON File and searches through the Firebase Data to determine what courses fit the constraints the user inputted. After generating a list of courses the website uses Twilio to message the user's phone number their recommended course list, so it can only be accessed by the user.

Challenges we ran into

Some challenges that we ran into were error checking the user input data so that it would account for the different variations of inputted text. For example some users may input CASCS111, while others may only input CS111. There is a huge inconsistency with user inputs, so we ran into difficulty reshaping that data into a singular format. Another difficult that ran into is the sheer amount of time that it took to parse through the data from the Course Search Tool website. Since it took about 30 minutes to fully parse through the data, we decided to send the parsed data to an external server so that we could make requests on that data instead of parsing through the website every-time we ran through the code.

Accomplishments that we're proud of

Since this is the first project we've ever submitted for a Hackathon we're proud of actually submitting a project that can be useful for all Boston University students struggling with course selection.

What we learned

We learned how to make our own API, deepen our knowledge of Amazon Web Services and what they can offer.

What's next for Digital Advisor

We hope to implement an actual registration function to the Digital Advisor so that its user doesn't need to manually register for the classes and also generalize our website so it can be used for other schools as well.