The inspiration for this project stemmed from the challenging experience of selecting schedules for many different courses here at UW-Madison. The process often lacked personalized filters to give way to individual preferences in terms of both average GPA and class timings. This led us to create a tool that would streamline this process, allowing students to control their scheduling and planning more than ever before!
-To create this web application, we needed to first obtain an accurate record of all Spring 2023 course offerings, including their meeting times, locations, professors, and more. This data was stored in a MongoDB which allowed for ease of use when web scraping (which was around 4 hour process in total) and retrieving data.
-We additionally had to obtain accurate grade averages for these classes to give students the ability to filter by classes with the lowest or highest GPAs in the past. We accessed MadGrades through web scraping and stored this additional data in our MongoDB.
-By using a Python Backend, we can create a comprehensive list of schedules for the class options provided, and then filter them based on the user's input preferences on the various factors.
-Once we determine the optimal schedule for the user, we create an image file as well as a bit of detailed text to return to the user through the Flask frontend.
-One of the first few challenges we ran into was the difficulty of scraping the data from the public course search and enroll website. These classes were not always listed consistently, sometimes with different terms and formatting used, as well as every class having different section counts and some lacked information on instructors or locations. This meant that we had to adapt our scripting scripts to adapt to these changing conditions and work with how the pages were presented.
-Another challenge we encountered was the integration of a webpage into a web hosting app, we had some experience hosting and developing static webpages, however, we had not created a dynamic webpage before. This meant that even though we had a working scheduling system, we struggled to properly host it.